Adoption of Libre Software in Public Institutions

Information Technology & People
Adoption of free/libre open source software in public organizations: factors of impact

Bruno Rossi Barbara Russo Giancarlo Succi
Article information:
To cite this document:
Bruno Rossi Barbara Russo Giancarlo Succi, (2012),"Adoption of free/libre open source software in public
organizations: factors of impact", Information Technology & People, Vol. 25 Iss 2 pp. 156 - 187
Permanent link to this document:
http://dx.doi.org/10.1108/09593841211232677
Downloaded on: 22 April 2016, At: 20:57 (PT)
Downloaded by Loyola University Chicago At 20:57 22 April 2016 (PT)
References: this document contains references to 55 other documents.

To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 1028 times since 2012*
Users who downloaded this article also downloaded:
(2010),"Open source software use in libraries", Library Review, Vol. 59 Iss 9 pp. 708-717 http://
dx.doi.org/10.1108/00242531011087033
(2012),"Open source deployment in local government: Rapid innovation as an occasion for
revitalizing organizational IT", Information Technology & People, Vol. 25 Iss 2 pp. 136-155 http://
dx.doi.org/10.1108/09593841211232631
(2005),"An introduction to the open source software issue", Library Hi Tech, Vol. 23 Iss 4 pp. 465-468 http://
dx.doi.org/10.1108/07378830510700678
Access to this document was granted through an Emerald subscription provided by emerald-srm:385160 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for
Authors service information about how to choose which publication to write for and submission guidelines
are available for all. Please visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company
manages a portfolio of more than 290 journals and over 2,350 books and book series volumes, as well as
providing an extensive range of online products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee
on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive
preservation.
*Related content and download information correct at time of download.

The current issue and full text archive of this journal is available at
www.emeraldinsight.com/0959-3845.htm
ITP
25,2 Adoption of free/libre open source
software in public organizations:
factors of impact
156
Bruno Rossi, Barbara Russo and Giancarlo Succi
Center for Applied Software Engineering (CASE),
Free University of Bolzano-Bozen, Bolzano, Italy
Abstract
Purpose – In this paper the authors aim to investigate the importance of factors for the adoption of
free/libre open source software (FLOSS) in the public sector. They seek to evaluate how different
factors impact during the initiation and implementation phases of the adoption process.
Design/methodology/approach – The authors base the methodological approach on two
exploratory case studies with a contrasting result logic. They build a multi-level framework
grounded both on literature review, and feedback from stakeholders. They then apply the framework
to two case studies to better frame the findings. They consider phases of adoption (initiation,
implementation) and the levels of adoption (technological, organizational, environmental, individual).
Findings – In the case studies, the authors found the importance of a strong and decision-centric
management board to give the impulse for the initiation phase of the process. As perceived by the
stakeholders, a strong governmental support is of paramount importance to increase the adoption at
the public level, although in the case studies examined the initiation stage started from the impulse of a
championing management. Both case studies passed the initiation phase successfully. Continuous
employees’ training, organizational objectives consensus, and business process reengineering have
been found important for the implementation phase. In the case study in which these factors were not
in place, the implementation phase of adoption failed. Environmental factors – although relevant for
the initiation of the adoption process – are less significant during the actual implementation of the
adoption process, as the contrasting result logic from the case studies shows.
Research limitations/implications – The study refers to two public organizations in a specific
environmental setting. No causality among factors has been inferred. Quantitative objective data have
been used to determine the success of adoption, for qualitative data multiple sources have been used
when possible to limit threats to validity.
Practical implications – The framework can be used by stakeholders in public organizations to
better frame their adoption strategies and to compare results across institutions. Lessons learnt from
the case studies can be useful to drive future adoptions of FLOSS.
Originality/value – The framework combines phases of adoption and levels making it possible to
frame the analysis of the case studies. It has been operationalized with a set of metrics, and with a
protocol for the case studies to increase replicability value.
Keywords Adoption, Diffusion theory, Case studies, Diffusion, Open source, Open source software,
Public sector organizations
Paper type Case study
This work has been partially supported by COSPA (Consortium for Open Source Software in the
Information Technology & People
Vol. 25 No. 2, 2012 Public Administration), EU IST FP6 Project No. 2002-2164. The authors would like to
pp. 156-187 acknowledge in particular the two public administrations that took part in the project.
q Emerald Group Publishing Limited
0959-3845
Acknowledgments in particular go to all the users involved, participants in the experimentation,
DOI 10.1108/09593841211232677 and technical personnel.
1. Introduction Adoption of
The advantages of Free/Libre Open Source Software (FLOSS) have been often FLOSS in public
discussed in recent information systems literature. Customizability, free access to the
source code, and costs savings are the major positive aspects reported by various organizations
research sources (Feller et al., 2005; Fitzgerald, 2005; Sanders, 1998). The benefits that
FLOSS can bring to the public sector are even more compelling, as adopting FLOSS
can reduce the digital divide phenomenon, promote local economies, avoid 157
vendor-dependence, and – in a broader view – attack the monopolistic dominance
on the market. These are all aspects that increase the importance of FLOSS in the
public sector.
Still, organizations in the public sector are reluctant to fully adopt FLOSS, since
triggers for the adoption in the public sector are different than for private
organizations. Very often, but not always, the public sector is driven by governmental
directives rather than single stakeholders’ inclination towards innovation (Huysmans

et al., 2008). In some cases, successful and unsuccessful FLOSS migrations were
reported even in case such driving force was missing (Välimäki et al., 2005; Ward and
Tao, 2009).
More specifically, why is the public sector somewhat reluctant to extensively adopt
FLOSS? A first reason concerns an endogenous aspect of innovation. Even
technologies that are beneficial to a community can take several years to be
extensively adopted, deployed, and used. Some anecdotal cases of problematic
introduction of innovations were the Dvorak keyboard (Margolis and Liebowitz, 1999),
the AC vs DC current (David and Bunn, 1988), laptop computers (Rogers, 1995), among
many others. In all these cases, switching costs and incremental returns in usage
played the major role in delaying the adoption process (Arthur, 1994).
There is another reason, though. FLOSS has been described as a disruptive
technology. While sustaining innovation improves existing values that customers
attribute to products, disruptive technologies revolutionize the way to think about a
product (Christensen, 1997). In this sense, three common characteristics of FLOSS are
typical of disruptive technologies (O’Reilly, 2005; Arina, 2005):
.
software as a commodity;
.
network-enabled collaboration; and
.
software customizability (software as a service).
First, the introduction of FLOSS poses the organization towards a potential radical
shift that considers FLOSS a commodity as public good driven by open standards and
open platforms. Then, a given class of FLOSS licenses eases the collaboration among
organizations giving them the chance to redistribute customized software through a
network. Finally, FLOSS allows a high degree of customization and opens the door to
new business frontiers. As such, the introduction of FLOSS can determine a significant
change in the organizational business processes, to adapt them to the new software.
This makes the adoption of FLOSS compelling, but also problematic: a migration to
FLOSS can involve a large deviation from the pre-existent organizational IT
infrastructure. This radical shift and the intrinsic mistrust toward innovation typically
slows down any FLOSS adoption. Thus, investigating factors of adoption facilitating
or inhibiting the introduction of a new technology is of primary importance in recent
literature on innovation adoption (Fitzgerald, 2009; Fitzgerald et al., 2011).
ITP In this paper, we follow the idea of a multilevel study for FLOSS adoption (Depietro
25,2 et al., 1990; Glynn et al., 2005) for which adoption is analyzed under different
perspectives and we identify factors that facilitate and inhibit FLOSS adoption in
public organizations. In this sense, our research specializes the general framework
introduced in (Fitzgerald et al., 2011) in that it focuses only on specific characteristics of
adoption, but it further analyses them at different phases of adoption and investigates
158 them in terms of a final measure of success. Specifically, our research question is:
What are relevant factors for a successful introduction of FLOSS at different phases of
adoption in a public organization?
To answer the question, we define a theoretical framework for FLOSS adoption inside
organizations, we refine the framework with stakeholders’ feedback, and then we apply
it to two exploratory case studies in public organizations. We contribute to the research
in FLOSS adoption with:

.
the definition of a multi-level and multi-phase framework for the characterization
of FLOSS adoption in the public sector;
.
the characterization of two large case studies, to understand the FLOSS adoption
phenomenon;
.
the evaluation in the case studies of the factors of FLOSS adoption by means of a
set of metrics identified, and the application of an empirical study protocol to
evaluate the intra-case study and inter-case studies results.
The article is structured as follows, in Section 2 we overview different studies that

identify factors of impact in FLOSS adoption. In Section 3 we propose the research
methodology and the process of framework building. In Section 4 we build the research
framework to study FLOSS adoption, drawing from the experience of other studies,
and we prioritize the factors by letting managers of two public organizations evaluate
the framework. In Sections 5, 6 we first define case studies used to draw common
lessons and then apply the framework to the two case studies. Discussions, limitations,
further works and conclusions end the article.
2. Background: factors of FLOSS adoption in literature

The theory of Innovation Diffusion Theory (Rogers, 1995) provides a basic context of
investigation for the adoption of new technologies that comprises several aspects, from
the types of adopters to the adoption phases. In this context, the factors of adoption can
be discussed at different levels or perspectives. At the personal/technological level,
studying motivations that act at the individual level and impact on the adoption
decision (among others, Dedrick and West (2004)), at the organizational level, studying
the factors inside organizations that act as facilitators/inhibitors for the adoption of
new technology (among others, Fichman, 2001), or at the environmental level, studying
the business dynamics and the impact of the external environmental factors on the
process of innovation adoption (among others, Economides and Katsamakas (2006)).
Larger frameworks have been built typically to integrate the three different levels in
the analysis of factors of a generic technological adoption: technological,
organizational, and environmental factors (Tornatzky and Fleischer, 1990; Swanson,
1994; Chau and Tam, 1997). Fitzgerald et al. (2011) have built a framework of FLOSS
adoption integrating Rogers’ (1995) facilitating conditions with Ajzen’s theory of Adoption of
Planned Behaviour (Gallivan, 2001; Ajzen, 1985; Taylor and Todd, 1995). FLOSS in public
The majority of the research in this area is performed from a single perspective.
Recent research focused more on multilevel studies (Jones and Gallivan, 2007). Such organizations
approach is greatly encouraged, though, as including different perspectives of analysis
allows studying a technology introduction in the organization as a whole. A multilevel
research would study interactions and similarities across various levels of adoption. In 159
our case, a multilevel approach considers all the four aspects simultaneously:
(1) individual;
(2) technological;
(3) organizational; and
(4) environmental.
Among the major contributions in factors of FLOSS adoption, Dedrick and West (2003)
developed a grounded theory of the adoption open source platform adoption, the case
of Linux Operating System. According to the authors, several factors have been
influencing FLOSS adoption decisions such as relative cost advantage perceived by
managers, compatibility with existing standards in the organization, trialability of the
new system, and complementary skills in the management of the organization. Two
environmental factors also played a major role, the perceived availability of FLOSS
skills in the external environment, and the support on FLOSS technologies by the
major IT vendors as external competences in FLOSS can foster the adoption and
organizations need not to rely solely on internal competences. The authors published a
second work (Dedrick and West, 2004), in which they administered to IT managers a
survey on the relevance of the factors they determined in their first work. With this
study they validated their previous work and classified the factors in three of the above
four aspects opening the way to a multilevel study. As a result, some of the factors
found previously were confirmed, such as relative cost advantage (technological
aspect) – also including switching costs (Shapiro and Varian, 1999). Some other factors
were added, such as reliability of the platform and compatibility with deployed
technologies and skills (technological aspect), IT Innovativeness as the general stance
of an organization toward technological innovation (organizational aspect), and
availability of external technological resources (environmental aspect).
Kwan and West (2005) examined FLOSS adoption in large enterprises. In particular,
they defined four critical attributes, industry context, firm context, organizational
attitudes towards standards, and towards open standard. The industry context is
considered critical, as it can influence the IT decisions of single companies. The firm
context also is determinant as different companies will have different competitive
positions and specific roles for the IT department. The organizational attitudes
towards standards and open standard are drivers of FLOSS adoption. The greater is
the attitude towards open data standards, the larger are the opportunities for the
adoption. The latter aspect is concerned in general with companies’ attitude towards
the FLOSS movement. Some companies are neutral in this aspect, while others are
unbalanced either towards proprietary software or FLOSS, increasing the probabilities
of a successful technological change inside the organization.
ITP The framework in Glynn et al. (2005) is the first multilevel study in FLOSS adoption,
25,2 which has been validated with a large-scale survey. As major relevant factors, the
authors included the FLOSS ideology, a committed personal championship of FLOSS,
and the network effects that can be exploited when collaborating with the large open
source community.
Other factors found relevant in complementary studies are the unfamiliarity of
160 customers with FLOSS vendors’ relationships, and the need to adapt corporate
governance and architecture (Holck et al., 2005). The complex interrelations between
multiple individual, technical, organizational, and environmental factors, including
total cost of ownership (Russo and Succi, 2009), external market support have also been
considered as relevant (Morgan and Finnegan, 2007).
The framework of Fitzgerald et al. (2011) integrates the principles of the theory of
Planned Behaviour adopted in Gallivan’s (2001) framework with Rogers’ facilitators of
innovation adoption. They analyzed the factors in terms of their effects on technology
assimilation. The framework is applied to five different case studies in a comparative
approach.
Following the initial seminal paper of Depietro et al. (1990) discussed recently in
Huysmans et al. (2008), factors have been divided in facilitators and barriers and
defined in the same way for the private and the public sector. This approach has been
used by many frameworks of FLOSS adoption hereinafter (Bradford and Florin, 2003;
Cooper and Zmud, 1990; Dedrick and West, 2003; Huysmans et al., 2008). Huysmans
et al. (2008) proposed factors introduced in Morgado et al. (2007) for the specific
adoption of FLOSS. In this work, the focus is more on the barriers that are relevant for
a specific FLOSS instance (e.g. OpenOffice.org) and at a finer level of detail (e.g. lack of
usability).
The tOSSad project identified several barriers to the adoption of FLOSS in the
public sector (Groganzt, 2007). Examples are human factors such as resistance to
change, trust on OS distribution, legal factors as software piracy, lack of business
models, and lack of FLOSS branding. Other factors that relate to technology are lack of
standardization of formats and compatibility, security problems and system failures,
versatility of the operating system installed. Additional barriers in public
administrations refer to the lack of a single reference for FLOSS selection (as in
proprietary software), and entry barriers for the migration process. Among those,
financial barriers are relevant. Crucial issues are difficulty of estimating the Total Cost
of Ownership of FLOSS and the organizational increase of costs such as the increase of
management costs, the new software compliance costs, etc. All of these factors have to
do with the difficulty of having an exact estimate of the resource needed for the
adoption and possible fear of unforeseeable expenses and hidden costs. To increase
uncertainty legal factors play a fundamental role. There is a lack of regulations in laws
about FLOSS use and adoption that relates to contracts and patent issues.
3. Research methodology
To answer our research question, we first build a theoretical framework of factors
facilitating/inhibiting technology adoption. We have defined it by following the
approach for framework definition of Eisenhardt (1989), and the multiple case study
design for research of Yin (2003). In this section, we present how we created the
framework. The framework itself will be described in the next section.
According to Eisenhardt (1989), the definition of a framework must follow several Adoption of
sequential steps. In the following, we describe how we implemented each single step of
the Eisenhardt’s method.
FLOSS in public
organizations
Getting started
In this step, we have defined the research question to build a multi-level, multi-phase
framework (Depietro et al., 1990; Huysmans et al., 2008). We considered both the 161
different level of each factor (technological, organizational, environmental, and
individual), and the phase on which they impact (initiation or implementation). We also
selected the relevant literature on factors facilitating or inhibiting technology adoption.
Selecting cases
We selected cases according to a multiple case study design for research (Yin, 2003). In
this kind of design, the procedure is replicated across case studies with different
context. It is more complex to be handled than a single case study design, but can
produce results that are more general. Specifically, we ran two case studies in two
different public administrations on which we used a contrasting result logic – or
theoretical replication, according to (Yin, 2003). The contrasting resulting logic
compares different results from the case studies to ground and refine theory and
strengthen findings. With this method, cases are not sampled, but are intentionally
selected to provide contrasting outcomes.
Crafting instruments and protocols

As means for data collection, we used surveys, interviews, pure observations, and
software agents. When feasible, we gathered data from multiple sources. With this
approach, we were able to collect both quantitative and qualitative data. Different means
and sources can help derive quality data for the analysis. To ensure the quality of the
replications, we also defined a case study protocol. Details are given in section 6.2.
Entering the field

Questions for interviews and questionnaires were refined iteratively. In particular, we
had to integrate questionnaires with interviews as new factors were discovered. This
falls into the principle of data collection adjustment (Eisenhardt, 1989), for which data
collection is not simply linear and sequential.
Analyzing data
According to theoretical replication method (Yin, 2003), we performed the analysis to
determine divergent outcomes from the case studies. Specifically, quantitative data
collected by software agents was compared by means of box plots whereas qualitative
data were interpreted according to the questions linked to the factors from questionnaires,
surveys, or observations. In Table I we show all the factors, the type of the data sources,
and the kind of question that was used to derive the factor as facilitator or inhibitor.
Shaping the hypotheses

Discussion of the findings was made according to tabular data provided by the framework
by means of the prioritization and the results of the instrumentation to the case studies.
We use two level of analysis. As a first level, we use the comparison of the prioritization
made at the management level against the factors evaluated as facilitators or inhibitors in
a case study. This analysis will show the understanding of the management of the factors
ITP
Metric Target
25,2
Initiation phase
Attitude towards Interviews/ Answer to the level of trust in the FLOSS solution that U/M
change Questionnaires was going to be installed. Questionnaire contained
question such as The substitution of the daily used
162 software with free software will cause for you: more
work in the short run, but advantages in the long run;
more work in the short run and no advantages in the
long run; less work; more work
Centralization Observation Number of managers that can impose the introduction M
of FLOSS. This has been evaluated by evaluating the
number of stakeholders to be influenced the
introduction. In some cases a single stakeholder, as a
unit director can impose the introduction to the whole

organization
Complexity Observation Structure of the management board in terms of M
managers that take the decisions. Has been evaluated
as the type of board that can take IT decisions
Formalization Interviews How formalized are the processes in the organization. M
Has been evaluated directly from the management
asking how they believe their processes have been
formalized in terms of procedures and protocols. The
context of public administrations sees typically highly
formalized protocols
Interconnectedness Observation/ Exchanges of information to external units for the M
Interviews sampled units. Has been evaluated by observing and
asking to stakeholders, how much the respondents
believe the units are interconnected
Organizational Interviews Potential financial resources available for the adoption M
slack process. Ha been evaluated informally by asking how
many resources are available for IT changes
Size Observation Size of the organization in terms of number of -
employees. This has been taken from statistics about
the organizations analysed, samples for the
quantitative results about the adoption are of the order
of the thousands
System openness Interviews Number of information exchanges with external M
organizations. Stakeholders were asked about the
current information flows to the external environment
Competitive Interviews Competition with other public administrations/ M
pressure organizations. Managers were interviewed about
possible competition with external entities
Governmental Interviews Existence of rules or norms that can trigger the M
rules adoption process. Stakeholders were asked about the
existence of rules that can ease the adoption process
Implementation phase
Technological Interviews Compatibility of FLOSS with the existing system M
compatibility infrastructure. Evaluation was carried out by
Table I. informally interviewing IT stakeholders about the
Method of evaluation of compatibility of OpenOffice.org in the context of their
the different factors infrastructure
selected (continued)
Metric Target
Adoption of
FLOSS in public
Technological Questionnaires Level of complexity of the software introduced. In this U
complexity moment, if Microsoft Office is removed, are you still organizations
able to perform the same tasks? Possible answers: yes;
yes but with some problems; no
Task complexity Interviews Peculiarity of the tasks to be performed with the usage U
of the new software. By means of informal interviews, 163
we evaluated the complexity of the tasks that need the
support of the new software
Relative advantage Questionnaires The perception that users have of the new technology. U
Evaluation was carried out by means of question: How
do you evaluate the functionalities of OpenOffice.org
with respect to the Microsoft Office ones? Possible
answers: widely superior; superior; equal; inferior;
widely inferior
Business processes Interviews Modification of existing business processes. Evaluation M
reengineering has been done by asking the IT management: have
been business processes inside the organization
reengineered to ease in some form the adoption of the
new software?
Top management Interviews Actions performed from the management to support M
support users during the adoption process. Evaluation was
done by asking the IT management: which kind of
support has been given for the introduction of the new
software?
Organizational Interviews Agreement of users with the objectives set by the top U
objectives management. This was evaluated by means of
consensus interviews with users about their specific agreement of
the decision to adopt OpenOffice.org. We asked
whether users agreed with the decision to introduce the
new software
Training Interviews Number and type of training activities performed. This M
information was gathered both by interviewing the IT
management
Note: Target can be either (M)anagers or (U)sers, or both Table I.
and their relevance in the context of the case studies. As a second level, we compare the
factors inside a single case study and between the two case studies, discussing the
divergences under the light of the different outcome of the case studies.
Enfolding literature
According to this step, the comparison with relevant similar and conflicting literature
increases internal and external validity of the study. We compared and discussed our
findings against literature in FLOSS adoption focusing on differences and similarities
from the determined.
Reaching closure
According to this step, more cases should not be added to the case when “incremental
learning is minimal” (Eisenhardt, 1989). This means that the introduction of new case
ITP studies will not give the theory an adequate benefit compared to the costs of carrying
25,2 out an additional investigation. We did not follow this step for two reasons: our aim is
to cross-check the factors across the case-studies, not to build a causal model.
Furthermore, due to the broad research question, the cardinality of factors, metrics, and
the size of the organizations adding more case studies would have been infeasible.
Figure 1 illustrates the tasks we achieved in building our framework with the
164 Eisenhardt’s approach. The first step is the definition of facilitating/inhibiting factors
for the adoption of FLOSS in public organizations. As we mentioned, we reviewed the
literature in the field to create a preliminary list. We started from the recent book of
Fitzgerald et al. (2011) and we went back reviewing factors from the major frameworks
of Innovation Diffusion Theory (Bradford and Florin, 2003; Dedrick and West, 2004;
Rogers, 1995; Cooper and Zmud, 1990). Each factor has been included according to the
following rules:
Figure 1.
Methodological approach
followed to build the
framework and integrate
the case studies
.
Not included already as a duplicate factor or subfactor. For example we Adoption of
considered software compatibility rather than the more fine-grained factor FLOSS in public
software compliance to standards. In general, we kept higher level factors.
.
Applicability during the subsequent case study research, based on the fact that
organizations
the factors could be collected in the case studies. For example we did not include
a factor such as level of software piracy in an organization, as we thought
difficult to let organizations of the case studies disclose this kind of information. 165
To build the framework, we mapped factors into the adoption phases (initiation and
implementation) as suggested by the Innovation Diffusion Theory (Rogers, 1995). Then,
we asked managers to review the resulting framework. Managers prioritized factors
according to their understanding of factors’ importance in the adoption phases. The final
framework is a distribution of prioritized factors across FLOSS adoption phases. To
apply our framework, we define a set of metrics that evaluate the factor of a given phase
(Table I). Metrics are specified with their type (qualitative or quantitative), the source
from which they were collected, and means of collection (surveys, interviews, pure
observation, or software agents). For example, we collected objective data on software
usage with software agents that collect time of use of software applications, we collected
subjective data on attitude toward change with direct questions or interviews to the
software users, and we collected objective data organization size by pure observations,
as suggested in Hannay and Jørgensen (2008), and in Sjøberg et al. (2005). Finally, we
applied the framework to two different case studies, measuring the factors and
comparing them in terms of similarities and contrasts.
In this context, we administered a questionnaire on the importance of the factors.
We explained the proposed framework, the different interactions, and terms, and asked
the respondents to prioritize the factors according to their experience and the two
phases of initiation and implementation.
4. FLOSS adoption integrated framework

We followed the research methodology explained in the previous section, and we
structured the framework as a multi-level and multi-phase model based on the general
innovation adoption framework of Rogers (1995) and the framework of OSS adoption
in (Fitzgerald et al., 2011). We describe in this section the framework.
The framework has two phases and four levels. The technology innovation process
in organizations has several interconnected phases (Rogers, 1995). In our framework,
we consider two of the Rogers’ macro-phases initiation and implementation – also
called primary and secondary adoption (Zaltman et al. (1973); Gallivan, 2001) – and we
do not delve into the single sub-phases. At the initiation phase, the organization takes
the decision to adopt a new technology, and at the implementation phase, the
technology is adopted inside the organization. Each phase is affected by different
factors that might inhibit or facilitate its development. For example, the effects of
governmental policies influence mostly the initiation phase as they impact on the
decision of the organization to adopt a particular technology. According to (Fitzgerald
et al., 2011), the four levels of our framework are technological, organizational,
environmental, and individual (TOEI).
The framework consists of factors that inhibit and facilitate FLOSS adoption
categorized by phases and levels define the mapping between phases and levels. For
ITP example, the factor “adverse to FLOSS ideology” can inhibits the introduction of
25,2 FLOSS during initiation at the organizational level as the new software application is
not considered appropriate to fit the identified needs or during implementation at the
individual level as users consider the application not compatible with previous legacy
formats.
The mapping is reported in Tables II and III. In the following, we discuss the
166 mapping by phase.
4.1 Initiation: environmental, individual organizational factors

In this section, we discuss levels that impact at initiation phase. In literature, mostly of
the factors that impact during the initiation phase are at the organizational,
environmental, and individual level.
Prioritized factors Description Source Score
1. Governmental rules Governmental rules that can (Groganzt, 2007; Hahn, 1.62
establish policies towards the 2002; Lee, 2006)
adoption of Open Source software
2. Attitude towards change Attitude towards change of the (Rogers, 1995) 1.69
management
3. Size Size of the organization (e.g. In (Rogers, 1995; Fichman and 4.0
terms of number of employees) Kemerer, 1997)
4. Complexity Heterogeneity of the knowledge (Rogers, 1995; Morgan and 4.23
of the management Finnegan, 2010)
5. Formalization How much formalized are the (Rogers, 1995) 5.46
processes inside the organization
6. Competitive pressure Level of competition with other (Rogers, 1995; Chau and 5.62
organizations (also other public Tam, 1997)
administrations)
7. Centralization Centralization of the decisional (Rogers, 1995; Dedrick and 5.92
power in a restricted number of West, 2003)
people inside the management
8. System openness The level of openness of the (Rogers, 1995) 6.0
organization towards the external
environment (e.g. with other
organizations)
9. Interconnectedness Level of connections between (Rogers, 1995) 6.15
organizational units and social
networks that can exist between
employees
10. Organizational slack Available resources not yet (Rogers, 1995; Dedrick and 9.08
assigned that can be devoted to West, 2003)
additional activities
Table II. Notes: Prioritization of selected factors, the lower the score the better, Intraclass correlation coefficient
Initiation phase (ICC) ¼ 0.53
Adoption of
Prioritized factors Description Source Score
FLOSS in public
1. Technological The compatibility of the new (Bradford and Florin, 2003; organizations
compatibility software with the old data Morgan and Finnegan, 2010)
formats and old legacy
applications 1.92
2. Training Training of the users to the new (Cooper and Zmud, 1990; Finney 167
software functionalities and Corbett, 2007) 2.31
3. Organizational Users Consensus with the (Cooper and Zmud, 1990)
objectives consensus objectives of the introduction of
the new software 2.38
4. Top management Indirect support of the (Cooper and Zmud, 1990; Glynn
support management to the users during et al., 2005; Morgan and
the deployment Finnegan, 2010) 2.92

5. Technological Complexity of the software to be (Bradford and Florin, 2003;
complexity introduced Morgan and Finnegan, 2010) 4.69
6. Business processes The necessity to modify the (Cooper and Zmud, 1990)
reengineering business processes inside the
organization to adapt them to the
new software 5.23
7. Task complexity Complexity of the tasks that the (Bradford and Florin, 2003)
employees have to perform 5.77
8. Relative advantage The perceived advantage that (Bradford and Florin, 2003;
users have when using a software Dedrick and West, 2003; Morgan
(e.g. better functionality, or and Finnegan, 2010)
performance, etc) 5.85
Table III.
Notes: Prioritization of selected factors, Intraclass correlation coefficient (ICC) ¼ 0.51 Implementation phase
At the organizational level, we consider factors of innovativeness of the organization,

defined as the degree of attitude toward the introduction of innovations in (Rogers,
1995). These factors are further classified as:
.
individual (leader) characteristics;
.
internal characteristics of organizational structure; and
.
external characteristics of the organization.
Each factor can impact positively (þ ) or negatively (2 ) in the three levels.

Centralization of the decisional power in few subjects inside the managerial board of
the organization has been found to be negatively correlated with the level of
organizational innovativeness as internal characteristic of an organization.
Complexity of the set of knowledge and expertise of management will positively
impact on the introduction of innovations. This means that a management that has
various competences will be more oriented towards innovations.
Formalization indicates the level of formalisms and bureaucracy that are in place
inside the organization. Higher formalisms in place will lead to greater barriers to the
introduction of innovations.
ITP Interconnectedness signals the level of connections among units in the
25,2 organizational social system. The more people are connected by interpersonal
relationships and social networks, the higher will be the spreading effect of innovations
inside the organizational setting.
Organizational slack is the availability of resources inside the organization not yet
assigned to specific activities. A larger availability of these resources will lead to a
168 greater attitude towards the introduction of an innovation, especially those that are
particularly resource-intensive in terms of implementation.
Size of the organization has been found to be positively correlated with the level of
organizational innovativeness. One reason can be that this indicator is in fact hiding
several other factors that have to do with organizational size, and those factors are
usually connected to innovativeness. Nevertheless, organizational size is easy to
measure and compare, and as such has been very often the basis for empirical studies
on technological innovation.
As an environmental characteristic, system openness defines how much the
organization is open to external influences in the evaluation of those innovations that
have been developed by external companies or that have been adopted with success by
equivalent organizations. Higher openness increases the knowledge about external
innovations that can be potentially beneficial. At the individual level, the attitude toward
change impacts positively on the organizational innovativeness: a management that is
change-oriented will favor the introduction of an innovation inside the organization.
Such management will easily see the potential benefits and impacts of innovations.
The influence of governmental rules, like strategies to support FLOSS diffusion, can
facilitate the organizational innovativeness in public organizations (Hahn, 2002; Lee,
2006) and be a facilitating factor for the adoption process.
Some of these factors have a double phase interpretation (Rogers, 1995). For example,
as we have seen, a high level of centralization leads to a lower degree of innovativeness in
the initiation phase, but the same level of centralization can be beneficial during the
implementation of the innovation, as such focused decisional power can better lead the
actual deployment of the innovation. This reason has been very often reported as one of
the causes for the scarce empirical use of this factor in a real scenario.
4.2 Implementation: individual, technical, environmental factors

We have seen that factors impacting on organizational innovativeness better refer to
the initiation of the adoption process; in this section we focus on the users’ acceptance
of innovation to determine those factors that impact on the implementation phase,
instead. In particular, we discuss the characteristics of Innovation Diffusion Theory
(Bradford and Florin, 2003; Dedrick and West, 2004; Rogers, 1995; Cooper and Zmud,
1990). In literature, these characteristics were not clearly related to the initiation and
implementation phases of Rogers.
The three major technological aspects discussed in (Bradford and Florin, 2003) refer
to:
(1) Technical compatibility, the level of alignment of the innovation with the legacy
technology.
(2) Technical complexity, the intrinsic difficulty in applying the innovation and
thus being successfully assimilated by users.
(3) Relative advantage, the advantage that users perceive that the innovation has Adoption of
with respect to old technology. FLOSS in public
We consider that the implementation of the innovation inside organizations is thus organizations
influenced by compatibility and technological complexity. Both task characteristics
and task complexity influence the alignment with compatibility and complexity as a
more complex task will potentially lead to greater compatibility issues. In this sense, 169
and intuitively, while the effect of increased technological compatibility is positive on
the final level of IT adoption, the effect of increased task and technological complexity
has clearly a negative impact on the adoption of innovation.
Technical compatibility refers thus to the level of fitting of the new innovation to the
existing software infrastructure. Both task characteristics and technology
characteristics determine the final level of compatibility. Bradford and Florin (2003)
show how technical compatibility can have a positive impact on the success of
deployment of a new software solution.
On the other side, technical complexity is defined as how an innovation is difficult to
understand and use (Rogers, 1995). Also in this case, the peculiarity of the task has a
final impact on the perceived usage complexity of a technology. In the framework
proposed, Bradford and Florin (2003) consider how the perceived degree of complexity
of IT systems will have a negative impact on success of deployment.
The reengineering of business processes is also considered an important factor in
implementation, as the introduction of a different software application may impose
changes in the organizations’ business processes to better fit the characteristics of the
new system (Bradford and Florin, 2003). In this sense, the introduction of a software
requires a radical change in the business processes of the organization so that the actual
implementation of the innovation can be slowed down or even hindered. In many cases,
reengineering business processes will have a positive impact on the organization, since
reengineering will optimize the process and reduce drastically wasted resources.
Bradford and Florin (2003) also included non-technical factors: the organizational
and environmental ones.
Organizational factors are constituted by four positive drivers: top management
support, organizational objectives consensus, and training, while the environmental
ones consists mainly of competitive pressure, for which organizations are confronted
with rival and equivalent organizations.
Top Management support refers to the vision and governance of the management
board in the technology adoption. As such, such management acts as facilitator.
Organizational objectives consensus, refers to employees that have clear
understanding of organizational objectives and are in agreement with them are
more motivated and inclines to the introduction of innovation.
Training refers generically to all the supporting and instructional activities
performed to decrease the barriers that users face when confronted with innovation.
Competitive pressure is an environmental factor (Porter, 1987; Bradford and Florin,
2003), that we do not consider impacting in the initiation phase. In our framework, we
considered it as a factor of initiation as, for example, adopting solutions already used
with success by rival organizations.
Figure 2 summarizes the mapping between levels and phases through factors. Figure 2
does not include all the factors we have found in literature. Many of these factors were not
ITP
25,2
170
Figure 2.
Integrated factors for
research in FLOSS
initiation and
implementation, according
to (T)echnological,
(O)rganizational,
(E)nvironment,
(I)ndividual levels
included either because they are sub-factors of other included factors or are too difficult to
operationalize in the data collection process. For example, factors such as the level of
pirated software inside an organization, the lack of branding, lack of business models, free
software perceived as “cheap” by users, were considered as difficult to operationalize in
the context of the case studies. We considered security problems and system failures, and
universality of the current operating system platform installed (Groganzt, 2007), more
related to a server-side installation than our focus on client applications.
Overall, the elements compounding our framework are consistent with the existing
frameworks we reviewed in Section 2 and the current research in innovation
introduction in the public sector (Raus et al., 2008). The framework does not introduce
new factors, but rather allows to discuss them to determine a theoretical mapping
TOEI levels – adoption phases that helps the application of the framework in concrete Adoption of
case of adoption (Figure 3). The novelty of our work will be evident in the application FLOSS in public
of the theoretical framework on two case studies (Table I) and in the prioritization of
the factors by the stakeholders. The final result – that will be presented after the organizations
discussion of the case studies – is a representation of the different factors as inhibitors
or facilitators by levels and phases in the two specific case studies (Figure 4). The
mapping allows to evaluate the results of a single case study (intra-case study) and 171
compare the factors across different case studies (inter-case study).
In the next sections, we describe the results derived from the two different case
studies considering both the intra- and inter-case studies analysis.
5. Case studies
The case studies are based on two public administrations that have migrated to FLOSS
in the last five years. In the following, we provide information about the relevant
characteristics of the two organizations and the study protocol that was applied to both
to operationalize the framework.
5.2 Suedtiroler Gemeinden Verein (SGV)

SGV is an Italian public association that provides services to a large set of townships.
The administration is a consortium of 116 municipalities, eight public offices
supplying social services, and about 30 offices. Nearly all the townships in the
consortium are small and have an average of 50 desktop machines, and most do not
have on-site technicians so maintenance is largely managed remotely from the
consortium headquarters. The budget available for ICT (Information and
Figure 3.
Integrated framework for
FLOSS adoption
investigation
ITP
25,2
172
Figure 4.
Application of the
framework to the two case
studies, with indication of
(F)acilitating and
(I)nhibiting factors
Communication Technology) services in such small municipalities is generally low.

The experience in FLOSS is relevant, acquired mainly server-side.
5.3 Autonome Provinz of Bozen-Bolzano (PROBZ)

The Autonome Provinz of Bozen-Bolzano is an Italian northern province that has a
so-called “Statuto Speciale”, and as such has many competences that usually are
assigned to the state administration. The administration of the province has a peculiar Adoption of
structure, due to the variety of activities that has to perform: on top of a general FLOSS in public
direction, there are 12 departments, down to 40 divisions, divided then in 178 offices.
While the inhabitants of the province are , 470,000, almost 25 percent of the organizations
inhabitants is employed in the Public Administration.
The IT division of the administration is composed of five offices that provide
technical services to all other divisions, from software customization, testing, and 173
deployment, to hardware support, installation, and substitution.
Study protocol. Once the organization decided to introduce FLOSS, we defined a
protocol to gather data from the various sources and with the various means described
above. The protocol defines the steps of data collection and sample definition. The
interested reader can refer to the work of Rossi et al. (2006), and Rossi et al. (2007) for
further details on the protocol’s implementation.
The first step has been the selection of the participants. Due to the number of
employees of the two organizations, we could not perform a full randomization of the
sample and employees were selected by the management. This exposed us to the threat
of a possible selection bias (Campbell and Stanley, 1966), as it could be that the most
change-oriented employees were selected. To mitigate this thread, we selected a large
sample of respondents in both the organizations: 1.486 (SGV) and 1.475 (PROBZ)
employees. Involving such a large number of employees and given the little knowledge
of the users of the FLOSS ideology, the change-oriented employees became a minority
and had a more limited impact on the results.
We submitted questionnaires to the users about their attitude towards FLOSS. This
was done to evaluate their knowledge about FLOSS, for example to define their
attitude with respect to FLOSS.
There were motivational seminars on the reasons of the experimentation that was
organized by each organization independently. Installation of the FLOSS solution,
OpenOffice.org, deployed by each technician in parallel with the old software. The
users could choose which application they preferred to use. We started at this step
monitoring the usage of the office suites for employees sampled. A software agent was
used to gather data in real-time from each client machine. We used this information to
evaluate later the successfulness of the migration.
During the course of the monitoring process, each organization organized training
sessions and support to the users. Due to the large number of employees to train, very
often organizations preferred to provide guidelines and support through help desks.
We submitted the final questionnaires and performed the interviews for the
investigation of the factors. Table I presents the final connections of questions from the
questionnaires to the factors.
6. Refinement and application of the framework

To refine and validate our framework we first prioritized the list of factors and then
applied the framework with the prioritized list to the two case studies.
We collected overall ten complete questionnaires from managers of public
organizations to prioritize the list of factors. Out of those, four were of managers of the
two case studies. Managers were either heads of IT departments or heads of specific
offices. As a ranking mechanism, we used the average of the position that was
identified for a factor by the respondents (the lower the value, the better). In Tables II
ITP and III, we provide the prioritized list of the factors, with the description, literature
25,2 source, and the average score given by the respondents and that we used for the overall
ranking. Some of the factors were selected more often than others as more important
(as for example Governmental rules, and attitude towards change, 1.62 vs 1.69, or
organizational slack that had a score of 9.08, meaning it was very often selected as the
least important factor). To determine the level of agreements among respondents, we
174 also calculated the Interclass Correlation Coefficient (ICC) of the rankings performed.
An ICC of 1.0 indicates a perfect agreement in that all the interviewees rank the factors
exactly the same. ICC has been calculated as a as a two-way mixed model, prescribing
that each question has been evaluated by the set of raters and that these raters were the
only ones of interest (Shrout and Fleiss, 1979). ICC calculated for both classes of factors,
shows similar and high agreement among the interviewers in both the cases of
initiation (0.53) and implementation (0.51).
Table II reports about the prioritization of factors for the initiation phase.
Governmental rules, attitude towards change, and size of the organization are
considered the top factors for the initiation process. Managers confirm the importance
of government support to the policies towards the introduction of FLOSS. Such policies
can be a strong factor for the adoption. Managers believed that without a strong
attitude towards change, it is difficult to get the initial impulse towards innovation.
Organizational size is seen as an important factor for the FLOSS migration. The
management has the perception that a larger organization can more easily undertake a
migration process.
System Openness, interconnectedness, and organizational slack are not seen as key
factors for a migration process. A greater openness towards the environment, a
stronger integration among the personnel of the organization, and the availability of
extra unused resources, are not seen as key factors.
Prioritization for the implementation level factors is presented in Table III. The top
factors for the implementation level are considered technological compatibility,
training, and organizational objectives consensus. Indeed, the compatibility with
legacy systems and data standards is important for any migration process. The
migration process must not pose serious issues of interoperability. Formal preparation
for the implementation phase is also considered important for the reduction of the
acceptance barriers. Another reason is that training is often seen as way to perform an
additional schooling of employees. The adoption is thus a beneficial excuse to enforce
this intention. A strong consensus towards the organizational objectives is also
considered a key factor: the acceptance of users will be greater in presence of an
alignment between their perception of the organizational objectives and their
agreement with those long term objectives.
Relative advantage, task complexity, and business process reengineering, are
considered less important for the implementation success. This means that the eventual
advantage – or disadvantage – of using FLOSS software, or even the complexity of the
tasks is not considered relevant: the organization can enforce the implementation. Even
the reengineering of the business processes is considered less relevant than other factors,
meaning that managers of the case studies consider the introduction of FLOSS not
strictly connected to a reorganization of the business processes.
Before applying our framework, we investigated the success of the adoption process
in the two phases in each case study. We determined success by using a software agent
to collect information about FLOSS usage in real-time. We found that the adoption in Adoption of
PROBZ reached lower levels of adoption. Data are illustrated in the box plots of FLOSS in public
Figure 5. In the specific, data for PROBZ show a distribution of usage of documents per
user has mean x; – ¼ 3.26 and standard deviation s ¼ 1:48. For SGV, the distribution organizations
of the usage of documents per user has mean x; – ¼ 12.38 and standard deviation
s ¼ 6:28. The time of usage per user yields similar results: in PROBZ the mean is x;
–
¼ 12.66 min. and the standard deviation is s ¼ 7:09, in SGV, x; – ¼ 115.5 min. with 175
s ¼ 66:5. We can see that FLOSS usage is higher for SGV in the implementation
phase. We also run the non-parametric Wilcoxon Rank Sum Test for the comparison of
the data distributions of the two case studies. Our null hypothesis is that both the
average documents and average time worked by users from the two case studies are
independent samples from identical continuous distributions with equal medians. In
both cases we can reject the null hypothesis with a 0.05 level of significance, two-tailed
(documents: p ¼ ,3,0392e-030 and time: p ¼ , 6.4333e-039). This result strengthens

the differences in the adoption levels.
Finally, we applied our framework to the two case studies collecting information on
phases, levels, and factors as perceived by the users. We collected the relevance of the
factors among a sample of volunteered users and the four managers of the two
organizations. The second column of Table I indicates the metric that was used to
derive the results for a factor in a specific case study. The third column of Table I
indicates whether a factor was evaluated by targeting a User (U) or a manager (M) as
the data source. Following Yin (2003), we further discuss the factors within the theory
of contrasting results in that case study design and protocol have been replicated in the
two case studies. We detail the findings in the following and in Tables IV-V, and
Figure 4.
6.1 Suedtiroler Gemeinden Verein (SGV)

The introduction of FLOSS in SGV was successful, as both the initiation and
implementation phases completed and the usage of the new technology was high. The
Figure 5.
Comparison of PROBZ
and SGV case studies,
boxplots about the
OpenOffice.org (OOo) final
usage compared
ITP
SGV
25,2
Initiation phase
Facilitators
Attitude towards change Informal interviews with the management, showed that there is a
strong evidence of attitude towards change from leaders in key
176 positions to drive the change. More sparse was the attitude of local
managers in the institutions and users
Centralization There is a high centralization of the power, a single person takes the
championship role to impose the introduction of the innovation, IT
managers have the decision control over packages to install
Complexity There are heterogeneous competences in the managing board that
allow the evaluation of the deployment of the new technology
Organizational slack According to the interviews with the management, resources are
available in quantity to grant the opportunity of the introduction of

FLOSS
Size The size of the organization is large, more than 4,000 users were
reported overall, the usage data in the case study considered 1,486
users
System Openness Managers report that the organization is in general open to other
administrations and to the citizens
Inhibitors
Formalization Interviews report that usually the processes inside the organization are
highly formalized, this imposed stricter constraints during the
migration
Interconnectedness The diffusion of FLOSS is spread across many municipalities,
communication is heavy only among employees in the same
municipality
Competitive Pressure Competitive pressure with other administrations is limited, in the sense
of competition in the Public Administration context, the competition is
heavy among public entities in the same area
Governmental rules Still there is not a clear governmental directive about the introduction
of FLOSS in public organizations, several managers reported this issue
Facilitators
Business processes Limited business process reengineering has been performed and has
reengineering been a facilitating factor for the introduction of the application as IT
managers reported
Task complexity In some cases, were the complexity was too high, the organization
performed an automation of the task, such as automatic generation of
certain documents
Organizational objectives There was generally a consensus (albeit not extended to the whole
consensus employees) about the introduction of the software
Top management support Support from managers was constant during the migration
Training Training was performed accurately and considered by users as very
important
Inhibitors
Technological Complexity Users were not completely satisfied with the user friendliness of the
application
Technological Compatibility The technology, while compatible with old formats had some issues
Table IV. with old formats
Framework applied to the Relative advantage Questionnaires with employees showed that, generally, they are neutral
SGV case study towards the new software
Adoption of
PROBZ
FLOSS in public
Initiation phase organizations
Facilitators
Centralization A single head of the IT department can give the impulse for the
introduction of FLOSS. The initiation phase can start from this
single impulse 177
Organizational slack Resources are available in quantity to grant the opportunity of
introduction of FLOSS
System openness The organization is in general open to other administrations and
to the citizens increasing opportunities for FLOSS introduction,
this was the starting cause for the initial impulse of introduction
Size The size of the organization is large in terms of employees, the
final usage data considered 1,475 users
Attitude towards change The attitude of management towards change is heterogeneous

among the different people
Complexity There are heterogeneous competences in the managing board
Inhibitors
Formalization Usually the processes inside the organization are, also in this
case, highly formalized, this imposed stricter constraints during
the migration
Interconnectedness The organization is spread across many buildings and offices,
communications is strong among employees in the same offices
Competitive pressure Competitive pressure with other administrations is limited
Governmental rules Still there is not a clear governmental directive about the
introduction of FLOSS in public organizations
Facilitators
Technological Compatibility The technology, while compatible with old formats had still some
issues with legacy documents
Top management support Support from managers was not so convinced after the initial
impulse from the championship, and the enthusiasm was fading
away
Task complexity There are some activities that require a great level of accuracy for
printed documents, in other the level of task complexity is low
Inhibitors
Relative advantage Interviews with the employees showed that employees do not see
advantageous the migration
Business processes reengineering No change to internal business processes
Technological complexity Users were not completely satisfied with the usage of the
application
Organizational objectives There was no clear consensus towards the introduction of
consensus software, employees were not widely informed about the
introduction of the application Table V.
Training Training was not performed accurately, mainly due to the Framework applied to the
difficulties in training large numbers of employees PROBZ case study
factors of the framework have been evaluated by the users and reported in Table IV,
where the positive and negative impact on FLOSS adoption are marked respectively as
(F)acilitators or (I)nhibitors, indicating the effect of each factor according to the metrics
identified (Table I).
ITP At the end of the implementation phase, many of the municipalities were using
25,2 OpenOffice.org in their activities as their main office automation platform. The
introduction of OpenOffice.org was made easier by the attitude towards change of the
management, willing to release the organization from vendor lock-in, to provide more
opportunities for local companies, and to reduce the cost of software licenses. This
attitude was particularly evident at the initiation level. By informal interviews with the
178 management, we discovered several roles that took a championing role towards
FLOSS, and gave the starting impulse towards the process of adoption. These roles
very often were in the key position to dictate the line of action in the IT department,
granting in this way a stronger impulse. Slack of resources were available for the
adoption process, although according to the prioritization step, these were not
considered a key factor for the initiation of the whole adoption process. We also
acquired knowledge from the stakeholders, that the openness of the organization for
the exchange of information granted an impulse for the introduction of FLOSS. This
was due to other organizations that were using different data standards. This was
another impulse for the management to start the adoption process.
Other factors were considered as sort of barriers to the initiation process, like the
formalization of the business processes inside the organization, or the limited
interconnectedness among organizational units that could help in spreading FLOSS
more rapidly. Governmental rules were not a key factor for the impulse towards the
adoption process. While stakeholders in the case study believe this is a key factor for
increasing the adoption of FLOSS, they agreed that in their case the impulse derived
mostly from the decision taken by a management that took the leading role to embrace
the change.
During the implementation phase, one of the most relevant factors in this case study
has been business process reengineering. The IT management took the opportunity to
integrate the new technology in their existing IT workflows so that open software
libraries could be used to generate documents in open standards. Users were then
forced to use the new technology to handle these documents, gaining then confidence in
the new technology.
To testimony the importance of compatibility, in one single municipality, that was
still handling legacy documents, the migration was cancelled, due to many
interoperability problems. This is the only case that has been reported of
non-adoption inside the whole organization. While the adoption was successful,
technological complexity was an inhibiting factor: sampled users reported in fact not to
be completely satisfied about the usability of the new software, and they did not find
relative advantages in using it during their daily work. Their decision to adopt the
software was mostly due to the modification of the workflows that imposed the usage
of the application and the support from the top management.
In this sense, the case study shows that the management successfully overcame the
negative factors at the implementation level that could potentially lead to a low users’
acceptance of the new software.
6.2 Autonome Provinz of Bozen-Bolzano (PROBZ)

As in the previous case study, the introduction of OpenOffice.org was made in parallel
with the old application: we analyze positive and negative factors for the initiation and
the implementation phase (Table V).
The introduction of FLOSS was unsuccessful on a large scale as it did not bring Adoption of
enough users to appreciate and use the new technology (Figure 5). The boxplots show FLOSS in public
the distribution of users in terms of average documents used during the period and
average time dedicated to the application. We can see a limited usage, both in terms of organizations
number of users that adopted the software, and in terms of the distribution of activity.
A comparison with SGV case shows that in PROBZ the final usage levels were largely
inferior (x; – ¼ 3.26 and s ¼ 1.48 versus x; – ¼ 12.66 and s ¼ 7.09). 179
Mostly, the initial installation of the application has been done in response to the
need to interoperate with other administrations. Apart from an initial support by the
management, the users were not pushed to adopt the new technology by means of
further actions such as informative meetings, or training sessions.
In the initiation phase, also in this case study we found the importance of the
management to start the adoption process. While in the previous case study the impulse
of adoption came directly from part of the management that took a championing role, in
this case study the adoption decision came from the need to interoperate with external
organizations during the exchange of information. Managers interviewed on this aspect,
confirmed that this was the main reason to start the whole adoption process. Also in this
organization, the availability of resources for the adoption process was consistent, but
not considered by managers as a discriminating factor to start the process.
Competitive pressure, governmental rules, interconnectedness of the organizational
units, and formalization were sort of barriers to the initial impulse towards the
initiation of the adoption process. Also in this case study, managers considered
governmental rules a strong facilitator to drive the adoption process, but they
recognized that in the case study this was not a factor that gave the impulse for the
adoption process.
In the implementation phase, we can get the importance of three factors that were
sort of inhibitors in this case study: support, organizational objectives consensus,
training, and business process reengineering. The initial support given to users was
reduced during the course of the migration, furthermore there was no real involvement
of users about the reasons of introduction of the technology. Where this has been done
results have been more encouraging (Rossi et al., 2006). Training has been performed
not convincingly and targeting a limited subset of users, due to the resources needed
and the limited impulse towards change of the management. Interviews with users and
managers confirmed that, indeed, training is considered by both as fundamental to
increase the acceptance of the technology. Especially when dealing with large number
of employees, it is easier to focus on other aspects of the migration, like customization
of the installation, communication to stakeholders, leaving aside training.
We must also consider that in this case study the impulse for the initial migration
came from the environment – the need to be aligned with other public institutions that
were pushing for the introduction of the new technology. Furthermore, the introduction
of FLOSS in this case was not a complete failure: there were many indirect benefits, like
a large deployment of the customized software for office automation on all clients of the
public organization, an initial evaluation of the benefits of the technology, and an
evaluation of the IT department about the opportunities of customization that
increased their competences for future further migrations. The next section discusses
in more detail the findings that can derived from Tables IV-V, and Figure 4 by
comparing the results from the two case studies.
ITP 7. Discussion
25,2 When we applied our framework to two real case studies, we aimed at understanding the
most relevant factors as perceived by the managers and users. We also wanted to
determine the factor effect on the adoption itself, as inhibitor or facilitator. The
instantiation of the framework in each case study has defined a model of adoption in the
corresponding organization. The two models have been discussed in the previous
180 sections, illustrated in Table IV-V, and compared in Figure 4. In particular, we found that
the majority of the factors are equally classified as inhibitors of facilitators in the two case
studies (Figure 4). This is certainly due to the similar context (same geographic region and
social environment and same economic sector) the organizations share. The interpretation
of the factors in the adoption process might slightly vary (e.g. “Attitude toward change”),
though. The minority of the factors that are classified differently in the two organizations
is the real novelty and can be associated to the success of the adoption. In particular, in
SGV case study, in which the FLOSS implementation succeeded, training, organizational
objective consensus, and business process re-engineering were relevant factors compared
to PROBZ case study. As such, we can hypothesize that training, consensus and process
re-engineering are crucial activities that drive the success of FLOSS adoption. Note that
some of the factors are very specific of FLOSS adoption. For example, organizational
objective consensus is a factor specifically related to the adoption of FLOSS as FLOSS is
not new software, but rather innovation as a new process of distribution and access to the
technology (Fitzgerald et al., 2011). Users might perceive no gain and benefit in migrating
to the same type of technology. In this sense, consensus might have no relevance in other
type of adoption. It is worth noticing that the top most relevant factors both in initiation
and implementation phase do not affect the success of the adoption. The most critical
factor that impair the success of a migration seems to be business process reengineering.
In the following, we respond to our primary research question with the findings of
the previous sections.
RQ. What are relevant factors for a successful introduction of FLOSS at different
phases of adoption in a public organization?
According to the perception of the management, some factors are more important than
others. Specifically, managers perceive relevant three factors: governmental rules that
push towards the introduction of FLOSS, the attitude towards change of the
management, and the size of the public organization as larger organizations will have
fewer barriers for the introduction of FLOSS (Fichman and Kemerer, 1997). The
relevance of size effect can be related to the fact that size is, in fact, a proxy for other
factors, like available resources for initiating an adoption process. Interesting, in this
sense, is that for managers the availability of unutilized internal resources
(organizational slack) is considered the least important factor for the initiation
phase. This finding is opposite to the one reported in (Dedrick and West, 2004;
Swanson, 1994) for which this factor cannot be underestimated. They state:
The relevance and impacts of slack resources in technology adoption has been a source of
contention in the literature. While slack is argued to provide the room needed for
experimentation, it is also argued that too much slack can reduce discipline and lead to
investment in pet projects with limited economic value (Swanson, 1994). The interesting point
in our findings is the fact that slack can take different forms (financial versus human
resources) with different impacts.
The result of reduced discipline in the investment of slack resources can be peculiar to Adoption of
the public sector, though, as the need to invest in experimentation is not always FLOSS in public
perceived important. If we look at the findings from the users, instead, this factor is
even considered a positive aspect of FLOSS adoption as resources are made available organizations
for any extra effort needed to the introduction of the technology.
Analyzing the implementation in the case studies, no governmental influence was
found relevant to start the adoption process. This was not in line with the prioritization 181
of the factors made by managers that considered the impulse of governmental
directives and rules of paramount importance for the introduction of FLOSS. Our
interpretation is that governmental rules are important to start the process of adoption,
and the management perceives them as important in the context of public
administrations. In the case studies examined, the importance de facto is more
limited than expected. An explanation is that impulse is important to start the process
of adoption, once such process has started, other factors become more important. This
is also the main difference with organizations in the private sector: the process of
initiation is more complex in the public sector due to larger external inputs that are
needed to begin the whole process – mostly governmental rules. Successively, factors
of impact during the implementation phases are similar to organizations that face the
introduction of FLOSS: technological, and organizational issues. This results is
conformant to the consideration of Hahn (2002) that a neutral government that should
not impose constraints on the adoption of a certain software typology, letting the
impulse come from the stakeholders, or to take into account the long term interests
from the society point of view (Lee, 2006). However, managers interviewed do not agree
with the view and would like a more proactive governmental impulse.
Similar to Morgan and Finnegan (2010), according to the management’s perception the
compatibility of the software to be introduced is relevant, and in our case it is even
considered the most important implementation factor. Furthermore, organizational
factors were perceived as more important during the implementation phase rather than
the initiation phase. In this sense, we found as in Morgan and Finnegan (2010) the
importance of top management support during the phases of the implementation process.
The analysis of the case studies data suggests that centralization, organizational
slack, size, organizational complexity, attitude towards change as the most positive
factors at the initial phase in both the organizations. There is a slightly different
interpretation of the factors though. The organization that also completed the
implementation phase had a management more prone to changes and to championing
FLOSS at the initial phase. The same organization considered the FLOSS migration an
opportunity to reengineer its business processes, automating tasks when possible to
allow a more transparent addition of the software inside the organization. Like in
Huysmans et al. (2008), we found the importance of considering the characteristics of
the organization during the implementation phase of FLOSS. This emerges by
comparing the two case studies (Figure 4): both organizations operate in the context of
a similar external environment, and governmental directives. The differences are
largely in the internal organizational support to the implementation phase of FLOSS. In
one case, there were no further actions to push the introduction of the new technology,
the adoption was merely left to the own willingness of the users. This finding is
consistent with the importance of a championing for the introduction of FLOSS in the
public administration, according to Glynn et al. (2005): “In terms of individual factors,
ITP the importance of support for OSS ideology and the existence of an OSS champion were
25,2 rated as most important. Again, as already mentioned this factor relates well to the
availability of OSS-literate personnel”. We found full confirmation of this finding in our
case studies.
8. Limitations
182 In this article, we did not investigate the factors of impact to infer the causality in their
relationships. Investigation of the causalities among factors implies building a richer
framework and investigating the different hypothesis of correlation and influences
between factors. We limited the analysis to the impact of the factors in the single case
studies.
About construct validity, we hampered possible concerns about the selection of the
outcome variable by means of the collection of evidence from multiple data sources, for
example by using questionnaires and interviews, reviewing discrepancies, and

contacting again participants. In cases where this was not possible, a single source of
evidence was used. After the data collection from the management by means of
questionnaires, the framework’s prioritization has been briefly discussed with them,
together with the instantiation of the framework.
For internal validity, the fact that no causal claims among the factors were made,
reduced biases in this category, in particular confounding threats, in which changes to
dependent variables can be due to another variable not under investigation. Still, we are
particularly subject to threats related to the instruments used for data collection and
involvement of researchers in the case studies. About the instruments, we submitted the
questionnaires electronically to avoid any influence of the submitters onto the answers of
participants and managers. When direct observations were involved, like when
determining the success of the different case studies by means of a software agent, we
postponed the time-window of analysis of several months to reduce Hawthorne effect.
Users knew they were monitored, but the longer time should have reduced the possible
bias of changing the type of activity due to the ongoing monitoring process. Due to the
qualitative nature of the research methodology, we are particularly subject to
experimenter bias. We believe that sample selection bias, and procedural bias were
limited. The protocol of the case studies was replicated with the aid of the IT workers of
the organizations without any aim to prove or disprove the successfulness of the
adoption. The usage of a multiple case study design with a theoretical replication (Yin,
2003), in which the contrasts of the different results from the case studies serve to ground
and refine theory and strengthen the findings hampered this problem. The two case
studies were selected among several that were performed, as such researchers had no
reason to report the successfulness of a particular case study. More severe concerns are
about measurement biases due to the qualitative nature of the factors’ measurement.
Given the aforementioned considerations about the instruments, we additionally
hampered this threat by letting each author review the material separately.
For external validity, the context of the Italian Public administration must be taken
into account. The specific FLOSS used has been OpenOffice.org, and the results have to
be considered under the light of this decision. Other type of software can lead to
different kind of findings, although we believe that for similar typology of software the
results in the case studies would have been equivalent. Results cannot be comparable
in cases of software that requires more radical changes, as an operating system.
9. Conclusions Adoption of
The introduction of FLOSS, although some undisputed advantages, is still problematic FLOSS in public
as FLOSS adoption reflects both the exogenous complexity of introducing a technology
and the endogenous complexity of disruptive innovation. organizations
To understand the impact of this complexity, we decided to create a framework that
evaluated factors influencing FLOSS adoption to have a greater control on what
determines the success of the adoption and use of a technology. We based our 183
framework on the major frameworks of the theory of diffusion of innovation (Bradford
and Florin, 2003; Dedrick and West, 2004; Rogers, 1995; Cooper and Zmud, 1990), on
literature review, and management prioritization following Eisenhardt (1989) steps.
We contributed to this theory identifying factors by the two phases of the adoption
process, initiation and implementation, and the levels of adoption, individual,
technological, organizational, and environmental (Rogers, 1995; Jones and Gallivan,
2007; Fitzgerald et al., 2011). We discussed factors in the specific case of FLOSS
introduction in public administrations refining the framework to capture all the aspects
relevant in our context. We submitted the list of factors to domain experts to define the
priorities of the factors. The result is a mapping between phases and levels through
factors prioritized and perceived by managers and/or users. The application of the
framework to two real case studies produced two models of FLOSS adoption that we
compared. We measured the success of an adoption phase with its completion and with
the level of assimilation and use of FLOSS.
The comparison between the models allows to understand which factors can be
involved in the success or failure of the adoption and which factors are in fact
perceived relevant by managers and/or users. From our study, we can draw some
recommendations to managers and users that aim at migrating to FLOSS:
. The cause of the success of a migration process need to be searched among
factors at the implementation phase.
.
Managers and users perceive benefits and shortcomings of a FLOSS migration
differently.
.
The failure of a migration may be better caused by organizational and individual
aspects.
.
The causes of failure or success might not be the ones perceived as most relevant
by managers at the beginning of a migration.
.
Managers should not underestimate the effect of training and users the effect of
individual consensus. Furthermore, a well-defined strategy of training can avoid
adverse attitude and motivate users and reaching users’ consensus to speed up
migration.
.
In addition, reaching consensus is a specific activity that needs to be performed
in a migration to FLOSS products as it is not the product that changes rather the
way the product can be manipulated and maintained.
The application of the framework to case studies by means of a contrasting result logic,
provided the way to evaluate the more important factors for FLOSS adoption in public
organizations, by evaluating the peculiarities of FLOSS as a disruptive technological
innovation. On one side, the framework allowed to see the differences among
managers’ perception of the importance of factors and the relevance of factors in the
ITP case studies. On the other side, the discrepancies among the two case studies. In the
25,2 case studies, it also allowed to evaluate how a comparable initiation phase cannot be
sufficient to reach the final adoption, other inhibiting factors can play their role to
hinder the whole adoption process.
References
184
Ajzen, I. (1985), “From intentions to actions: a theory of planned behaviour”, in Kuhl, J. and
Beckmann, J. (Eds), Action Control: From Cognition to Behaviour, Springer, New York,
NY, pp. 11-39.
Arina, T. (2005), “The state of FLOSS and future opportunities”, available at: http://flosse.dicole.
org/media/documents/ArinaState_of_FLOSS_and_Future_Opportunities-Draft.pdf
(accessed December 3, 2008).
Arthur, W.B. (1994), Increasing Returns and Path Dependency in Economy, University of
Michigan Press, Ann Arbor, MI.
Bradford, M. and Florin, J. (2003), “Examining the role of innovation diffusion factors on the
implementation success of enterprise resource planning systems”, Information Systems,
Vol. 4, pp. 205-25.
Campbell, D.T. and Stanley, J.C. (1966), Experimental and Quasi-experimental Designs for
Research, Rand-McNally, Chicago, IL.
Chau, P.Y.K. and Tam, K.Y. (1997), “Factors affecting the adoption of open systems:
an exploratory study”, MIS Quarterly, Vol. 21 No. 1, pp. 1-24.
Christensen, C.M. (1997), The Innovator’s Dilemma: When New Technologies Cause Great Firms
to Fail, Harvard Business School Press, Boston, MA.
Cooper, R.B. and Zmud, R.W. (1990), “Information technology implementation research:
a technological diffusion approach”, Management Science, Vol. 36 No. 2, pp. 123-39.
David, P.A. and Bunn, J.A. (1988), “The economics of gateway technologies and network
evolution: lessons from electricity supply history”, Information Economics and Policy,
Vol. 3, pp. 165-202.
Dedrick, J. and West, J. (2003), “Why firms adopt open source platforms: a grounded theory of
innovation and standards adoption”, Standard Making: A Critical Research Frontier for
Information Systems Proceedings, Seattle, WA, pp. 236-57.
Dedrick, J. and West, J. (2004), “An exploratory study into open source platform adoption”,
Proceedings of the 37th Annual Hawaii International Conference on System Sciences
(HICSS ’04). Track 8 Vol. 8, IEEE Computer Society, Washington, DC.
Depietro, R., Wiarda, E. and Fleischer, M. (1990), “The context for change: organization,
technology and environment”, in Tornatzky, L.G. and Fleischer, M. (Eds), The Processes of
Technological Innovation, Lexington Books, Lexington, MA, pp. 151-75.
Economides, N. and Katsamakas, E. (2006), “Two-sided competition of proprietary vs open
source technology platforms and the implications for the software industry”, Management
Science, Vol. 52 No. 7, pp. 1057-71.
Eisenhardt, K.M. (1989), “Building theories from case study research”, Academy of Management
Review, Vol. 14 No. 4, pp. S532-S550.
Feller, J., Fitzgerald, B., Hissam, S. and Lakhani, K. (2005), Perspectives on Free and Open Source
Software, MIT Press, Cambridge, MA, pp. 121-40.
Fichman, R.G. (2001), “The role of aggregation in the measurement of IT-related organizational
innovation”, MIS Quarterly, Vol. 25 No. 4, pp. 427-55.
Fichman, R.G. and Kemerer, C.F. (1997), “The assimilation of software process innovations: Adoption of
an organizational learning perspective”, Management Science, Vol. 43 No. 10, pp. 1345-63.
Finney, S. and Corbett, M. (2007), “ERP implementation: a compilation and analysis of critical
FLOSS in public
success factors”, Business Process Management Journal, Vol. 13 No. 3, pp. 329-47. organizations
Fitzgerald, B. (2005), “Has open source a future?”, in Feller, J., Fitzgerald, B., Hissam, S. and
Lakhani, K. (Eds), Perspectives on Free and Open Source Software, MIT Press, Cambridge,
MA, pp. 121-40. 185
Fitzgerald, B. (2009), “Open source software adoption: anatomy of success and failure”, IJOSSP,
Vol. 1 No. 1, pp. 1-23.
Fitzgerald, B., Kesan, J.P., Russo, B., Shaikh, M. and Succi, G. (2011), Adopting Open Source
Software: A Practical Guide, The MIT Press, Cambridge, MA.
Gallivan, M.J. (2001), “Organizational adoption and assimilation of complex technological
innovations: development and application of a new framework”, ACM SIGMIS Database,
Vol. 32 No. 3, pp. 51-85.

Glynn, E., Fitzgerald, B. and Exton, C. (2005), “Commercial adoption of open source software:
an empirical study”, Empirical Software Engineering, International Symposium
Proceedings. IEEE Computer Society, Washington, DC.
Groganzt, G. (2007), “tOSSad Project. Obstacles and Barriers to F/OSS Adoption in Public
Administration. Annex I”, available at: www.tossad.org/content/download/1386/6897/file/
tOSSad_D18_Annex1_V2.3.pdf (accessed September 2, 2010).
Hahn, R.W. (2002), “Government policy toward open source software: an overview”, in Hahn, R.W.
(Ed.), Government Policy toward Open Source Software, Brookings Institution Press,
Washington, DC, pp. 12-33.
Hannay, J. and Jørgensen, M. (2008), “The role of deliberate artificial design elements in software
engineering experiments”, IEEE Transactions in Software Engineering, Vol. 34 No. 2,
pp. 242-59.
Holck, J., Larsen, M.H. and Pedersen, M.K. (2005), “Managerial and technical barriers to the
adoption of open source software”, COTS-Based Software System, pp. 289-300.
Huysmans, P., Ven, K. and Verelst, J. (2008), “Reasons for the non-adoption of OpenOffice.org in a
data-intensive public administration”, First Monday, Vol. 13 No. 10.
Jones, A. and Gallivan, M. (2007), “Toward a deeper understanding of system usage in
organizations: a multilevel perspective”, MIS Quarterly, Vol. 31 No. 4.
Kwan, S.K. and West, J. (2005), “A conceptual model for enterprise adoption of open source
software”, in Bolin, S. (Ed.), The Standards Edge: Open Season, Sheridan Books, Ann
Arbor, MI.
Lee, J.A. (2006), “Government policy toward open source software: the puzzles of neutrality and
competition”, Knowledge, Technology, and Policy, Vol. 18 No. 4, pp. 113-41.
Margolis, S.E. and Liebowitz, S. (1999), Winners, Losers & Microsoft: Competition and Antitrust
in High Technology, The Independent Institute, Oakland, CA.
Morgan, L. and Finnegan, P. (2007), “How perceptions of open source software influence
adoption: an exploratory study”, Proceedings of the 15th European Conference on
Information Systems (ECIS 07), pp. 973-84.
Morgan, L. and Finnegan, P. (2010), “Open innovation in secondary software firms: an exploration
of managers’ perceptions of open source software”, DATA BASE for Advances in
Information Systems, Vol. 41 No. 1, pp. 76-95.
Morgado, G., van Leeuwen, M., Özel, B. and Erkan, K. (2007), “Current status of F/OSS”, tOSSad
Deliverable D18, March 18.
ITP O’Reilly, T. (2005), “The open source paradigm shift”, in Cusumano, M., Shirky, C., Feller, J.,
Fitzgerald, B., Hissam, S.A. and Lakhani, K.R. (Eds), Perspectives on Free and Open Source
25,2 Software, The MIT Press, Cambridge, MA.
Porter, M.E. (1987), From Competitive Advantage to Corporate Strategy, Harvard Business
Publishing, Boston, MA.
Raus, M., Kipp, A. and Boutellier, R. (2008), “Diffusion of e-government IT innovation: a case of
186 failure?”, Proceedings of the eChallenges 2008 Conference, Stockholm, October 22-24,
pp. 240-7.
Rogers, E.M. (1995), Diffusion of Innovations, 4th ed., Free Press, New York, NY.
Rossi, B., Russo, B. and Succi, G. (2007), “Evaluation of a migration to open source software”, in St
Amant, K. and Still, B. (Eds), Handbook of Research on Open Source Software:
Technological, Economic, and Social Perspectives, Idea Group Reference, Hershey, PA.
Rossi, B., Scotto, M., Sillitti, A. and Succi, G. (2006), “An empirical study on the migration to
OpenOffice.org in the public administration”, International Journal of Information

Technology and Web Engineering, Vol. 1 No. 3, pp. 64-80.
Russo, B. and Succi, G. (2009), “A cost model of open source software adoption”, IJOSSP, Vol. 1 No. 3.
Sanders, J. (1998), “Linux, open source, and software’s future”, IEEE Software,
September/October, pp. 88-91.
Shapiro, C. and Varian, H.R. (1999), Information Rules: A Strategic Guide to the Network
Economy, Harvard Business Press, Boston, MA.
Shrout, P.E. and Fleiss, J.L. (1979), “Intraclass correlations: uses in assessing rater reliability”,
Psychological Bulletin, Vol. 86 No. 2, pp. 420-8.
Sjøberg, D.I.K., Hannay, J.E., Hansen, O., Kampenes, V.B., Karahasanovic, A., Liborg, N.-K. and
Rekdal, A.C. (2005), “A survey of controlled experiments in software engineering”, IEEE
Transactions on Software Engineering, Vol. 31 No. 9, pp. 733-53.
Swanson, E.B. (1994), “Information systems innovation among organizations”, Management
Science, Vol. 40 No. 9, pp. 1069-92.
Taylor, S. and Todd, P. (1995), “Understanding IT usage: a test of competing models”,
Information Systems Research, Vol. 6 No. 2, pp. 144-76.
Tornatzky, L.G. and Fleischer, M. (1990), The Processes of Technological Innovation, Lexington
Books, Lexington, MA.
Välimäki, M., Oksanen, V. and Laine, J. (2005), “An empirical look at the problems of open source
adoption in Finnish municipalities”, Proceedings of the 7th International Conference on
Electronic Commerce (ICEC ’05), Vol. 113, ACM, New York, NY, pp. 514-20.
Ward, D. and Tao, E.Y. (2009), “Open source software use in municipal government: is full
immersion possible?”, Proceedings of The World Congress on Engineering and Computer
Science 2009, pp. 1044-9.
Yin, R.K. (2003), Case Study Research. Design and Methods, 3rd ed., Sage, London.
Zaltman, G., Duncan, R. and Holbeck, J. (1973), Innovations & Organizations, Wiley & Sons, New
York, NY.
Further reading
Janssen, M., Weerakkody, V. and Dwivedi, Y.K. (2009), Handbook of Research on ICT-Enabled
Transformational Government: A Global Perspective, IGI Global, Hershey, PA.
Shadish, W.R., Cook, T.D. and Campbell, D.T. (2001), Experimental and Quasi-Experimental
Designs for Generalized Causal Inference, 2nd ed., Houghton Mifflin Company, Boston, MA.
About the authors Adoption of
Bruno Rossi is a non-tenured researcher at the Faculty of Computer Science, Free University of
Bozen-Bolzano. He holds a BSc degree in Applied Computer Science, and an MSc degree in FLOSS in public
Economics. In April 2008, he received the PhD degree in Computer Science from the Free organizations
University of Bolzano-Bozen. His research interests are in the areas of software engineering that
deal with free/libre open source software development models and software evolution. In the past
few years, he has been appointed as teacher of software evolution and requirements and design
of software systems courses. He has participated in important European (COSPA, STREP FP6) 187
and Italian projects (ArtDeco, FIRB 36 months), and has been involved as an active member or
reviewer in several conference program committees (OSS, HICSS, ESEM, IJCNN, NaBIC). Bruno
Rossi is the corresponding author and can be contacted at: bruno.rossi@unibz.it
Barbara Russo received her PhD degree in Mathematics from the University of Trento, Italy,
in 1996. She was visiting researcher at the Max-Planck-Institut für Mathematik in Bonn (D) and
at the University of Liverpool (UK). She is currently Associate Professor at the Faculty of
Computer Science of the Free University of Bozen-Bolzano, Italy. Her current research interest
focuses on various topics of software engineering specifically related to software metrics,

software reliability, data analysis and prediction systems with particular focus on open source
software and agile development. She has participated in the FP5 EU project NAME and the FP6
EU project COSPA and TEKNE. She is coordinator of the European Master Course in Software
Engineering (EMSE), the joint international Master course awarded as Erasmus Mundus. She is
Director of the Council of the Master at the Free University of Bozen-Bolzano. She is a reviewer of
several journals (e.g. Journal of System and Software and Empirical Software Engineering
Journal ) and conference program committees (e.g. Open Source Software and System (OSS) and
Empirical Software Engineering and Measurement (ESEM)).
Giancarlo Succi is Professor with Tenure at the Free University of Bozen-Bolzano, Italy,
where he directs the Centre for Applied Software Engineering and is the Dean of the Faculty of
Computer Science. Before joining the Free University of Bozen-Bolzano, he has been Professor
with Tenure at the University of Alberta, Edmonton, Alberta, Associate Professor at the
University of Calgary, Alberta, and Assistant Professor at the University of Trento, Italy. The
research interest of Giancarlo Succi involve multiple areas of software engineering, including
open source development, agile methodologies, experimental software engineering, software
engineering over the internet, and software product lines and software reuse. Giancarlo Succi is a
Fulbright Scholar. He coordinated the FP5 EU project Name and the FP6 EU project COSPA. He
is a member of ISERN, the International Software Engineering Research Network, isern.iese.de,
coordinated by the Fraunhofer Institute in Software Engineering.
To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints
This article has been cited by:
1. Maha Shaikh. 2016. Negotiating open source software adoption in the UK public sector. Government
Information Quarterly 33, 115-132. [CrossRef]
2. Mpho Ngoepe Department of Information Science, University of South Africa, Pretoria, South Africa .
2015. Deployment of open source electronic content management software in national government
departments in South Africa. Journal of Science and Technology Policy Management 6:3, 190-205.
[Abstract] [Full Text] [PDF]
3. Lakshmanan Ramanathan Birla Institute of Technology and Science (BITS), Pilani, India Sundaresan
Krishnan Infosys, Mysore, India . 2015. An empirical investigation into the adoption of open source
software in Information Technology outsourcing organizations. Journal of Systems and Information
Technology 17:2, 167-192. [Abstract] [Full Text] [PDF]

Adoption of Libre Software in Public Institutions

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Adoption of Libre Software in Public Institutions

Uploaded by

Copyright:

Available Formats

Information Technology & People

Adoption of free/libre open source software in public organizations: factors of impact

References: this document contains references to 55 other documents.

*Related content and download information correct at time of download.

directives rather than single stakeholders’ inclination towards innovation (Huysmans

in FLOSS adoption with:

The article is structured as follows, in Section 2 we overview different studies that

2. Background: factors of FLOSS adoption in literature

Crafting instruments and protocols

Entering the field

Shaping the hypotheses

unit director can impose the introduction to the whole

Note: Target can be either (M)anagers or (U)sers, or both Table I.

4. FLOSS adoption integrated framework

4.1 Initiation: environmental, individual organizational factors

Prioritized factors Description Source Score

the deployment Finnegan, 2010) 2.92

At the organizational level, we consider factors of innovativeness of the organization,

Each factor can impact positively (þ ) or negatively (2 ) in the three levels.

4.2 Implementation: individual, technical, environmental factors

5.2 Suedtiroler Gemeinden Verein (SGV)

Communication Technology) services in such small municipalities is generally low.

5.3 Autonome Provinz of Bozen-Bolzano (PROBZ)

6. Refinement and application of the framework

(documents: p ¼ ,3,0392e-030 and time: p ¼ , 6.4333e-039). This result strengthens

6.1 Suedtiroler Gemeinden Verein (SGV)

available in quantity to grant the opportunity of the introduction of

Attitude towards change The attitude of management towards change is heterogeneous

6.2 Autonome Provinz of Bozen-Bolzano (PROBZ)

example by using questionnaires and interviews, reviewing discrepancies, and

Vol. 32 No. 3, pp. 51-85.

OpenOffice.org in the public administration”, International Journal of Information

focuses on various topics of software engineering specifically related to software metrics,

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

You might also like