Competing for access and re-use? : retro-digitization partnerships in Dutch academic libraries.

(1)

COMPETING FOR ACCESS AND RE-USE?

Retro-Digitization Partnerships in Dutch Academic Libraries

Brigit Hoomans, 10278702

Master Thesis, Final version 31 January 2014

Amsterdam Business School

MSc. Business Studies, Strategy Track

(2)

1. ABSTRACT

“Make everything available”. With Anderson’s (2008) maxim in mind, academic libraries enter into partnerships with publishing entities in order to finance retro-digitization efforts they otherwise cannot afford. The question addressed in the present research is if these efforts, from a business perspective, contribute to the academic libraries’ strategic goal to provide online access to their information resources as well as to stimulate their re-use, and thus if these efforts lead to a stronger competitive position. Particularly relevant are the differing public and private perspectives on online access and re-use. In order to create sustainable advantage and to protect his investments, a publisher will wish to create barriers to

competition, such as exclusivity terms and property rights. Accepting such conditions, however, will to a large extent determine the library’s opportunities to create free access to their user communities, and consequently influence their performance.

The efforts of academic libraries were examined in three cases. Results to a large extent confirm previous research that public organizations such as libraries often lack the required capabilities to optimize project cooperation and results. In this specific context, libraries do not meet all the recommendations found in the literature with regard to exclusivity, term, and property rights of digitized materials. As a result of the contractual clauses agreed on,

academic libraries find they cannot facilitate new kinds of demand from both existing and new user communities.

Another finding is that libraries thus far have not implemented specific performance indicators to measure the effects of retro-digitization on online access and re-use, despite detailed suggestions in the literature. It proved difficult, therefore, to draw underpinned conclusions regarding their competitive position, other than the above-mentioned negative impact on new business opportunities.

(4)

2. INTRODUCTION

“Make everything available”. Although Chris Anderson’s long tail business model (Anderson, 2008) was particularly describing the new business opportunities of online retailing, his Rule Number 1 shook up strategists in quite another environment as well. His words echoed in archives and libraries worldwide, where “everything” already was available – though not yet digitally. These institutions were familiar with retro-digitization of their collections by then. Government funded national digitization programmes had originally focused on preserving fragile materials only, though. With the long tail in mind, subsidies now were asked for, and granted, to programmes that particularly aimed at the creation of online access and re-use of cultural materials to existing and potentially interested new groups of users. Dutch examples of such programmes are Beelden voor de Toekomst (Images for the Future), the national

programme to digitize audio-visual collections; or Metamorfoze, the national programme to digitize Dutch newspapers, Dutch books and manuscripts.

Businesses around the world at the same time started looking for more online content,

including multinationals such as telecom providers, Microsoft, and Google. Publishing entities such as Kluwer, Springer and Elsevier started their own retro-digitization programmes, as e-business for books and journals in the academic publishing market began to grow.

The economic perspective has changed since 2008, however. Government funding for retro-digitization, because of the recession, declined, or stopped altogether. And even though a substantial part of cultural collections in the meantime was made available online, the work done was still the top of the iceberg only. From an academic library perspective, therefore, it became necessary to find additional financial resources in order to keep up their digitization efforts (Nicholas et al., 2010; Kaiser, 2012; Savenije et al. 2012; Leonardo, 2012). Entering into partnerships with private parties such as publishers proved a way to do so. Publishing entities interested in retro-digitization projects are mainly aggregating publishers, traditionally

(5)

operating in the academic market segment only (Bell, 2012). Digital technology facilitated the development of new publishing models, because of cost reduction in the production processes on the one hand, and enabling an increase in access on the other (Brown et al., 2007).

Typically, these publishing entities partner with more than one academic library in order to create their online resource portfolio. Examples are Sage, Bloomsbury, ProQuest, and Brill. A notable outsider is Google, that started its Book Search Programme in 2004, with the aim to make millions of printed books accessible online, free of charge, to users worldwide (Sandler, 2005).

Partnering in retro-digitization for online access and re-use purposes, therefore, seemed to be an opportunity to create synergy and sustainable competitive advantage to both parties. In this thesis, three examples of such partnerships in the field of digital humanities are examined with the aim to investigate if and how especially the academic library partners are able to use the project results to strengthen their competitive positions. Particularly interesting are the differences in outlook on online access and re-use, which from a library’s – public – perspective is open and free of charge to as many users as possible; whereas publishers generally prefer a business model that regulates access through purchase or subscription. Moreover, to create sustainable advantage and to protect his investments, the publisher will wish to create barriers to competition (Porter, 1979), such as contractual clauses on exclusivity and property rights. Accepting such conditions, however, will to a large extent determine the library’s opportunities to create free access to, and meet the demands of, their existing and new user communities, and consequently influence their performance.

This thesis starts with a review of relevant literature on the topics described above. The Design and Method Section explains the way in which the research was conducted. The results of the research are presented in the Findings Section, and discussed in the Discussion Section. In

(6)

this section, several suggestions for further research are included as well. Finally, in the Conclusion, scientific and managerial implications of this research are presented.

(7)

3. LITERATURE REVIEW

This section starts with a short description of the role and strategy of academic libraries, together with resources and performance indicators related to retro-digitization efforts and online access. Regarding their digital strategy, academic libraries need to solve a financial deficit as the larger part of their budget is still appropriated to physical information resources. From their perspective, finding additional funding in retro-digitization partnerships with private parties is a way to create value. Detailed attention is paid to the topics that academic libraries need to negotiate when entering into this particular kind of partnerships, i.e. exclusivity, term and property rights of digitized information resources, as these factors determine the libraries’ opportunities to create online access and to stimulate the re-use of their content – which is the reason to start retro-digitization programmes in the first place.

3.1 ACADEMIC LIBRARIES’ ROLE, STRATEGY AND RESOURCES

Generally speaking, the role of academic libraries, as centers of knowledge dissemination and learning, is to collect, preserve and distribute societal intellectual output in order to enhance the research and educational missions of their parent universities (Heidorn, 2011). Academic libraries for their user communities traditionally serve as a gateway for discovering and using information resources such as books, journals or databases, as a buyer of these information resources, and as a repository of these resources (Schonfeld & Housewright, 2010).

In the digital age, academic libraries defined the creation of online access to their information resources as a strategic goal, due to changing behaviour of library user communities that more and more prefer online access to content (Brown et al. 2011), together with financial

considerations such as increasing space and storage cost of physical collections, and increasing purchase costs (Hodges et al., 2010).

(8)

From the resource-based perspective, in order to create sustainable competitive advantage academic libraries should consider their valuable, rare, inimitable and non-substitutable resources (Barney, 1991; Prahalad & Hamel, 1990). One of those VRIN resources is their physical collection, often containing age-old unique content. As library users not only expect online access to “born digital” information, but the more and more demand online access to older content as well (Brown et al., 2011), retro-digitization of unique backlist resources for online access and re-use purposes is a major opportunity to create value (Hodges et al., 2010; Marks & Janke, 2012).

Regarding resources, however, in order to reach this strategic goal academic libraries need to address two major issues. First, they need financial resources as retro-digitization projects come with steep upfront investments for digitizing, and large fixed costs for creating and

maintaining an IT-infrastructure for long-term access to their digitized content (Müller-Langer et al., 2011). From especially the perspective of the libraries this is a serious bottleneck, as their budget usually does not include digital services (Hodges et al., 2010). Internal library budgets still for the larger part are appropriated to providing access to physical collections, including for instance cost for long-term preservation, storage, and access (Odlyzko, 1999). Moreover, libraries are facing a decrease in their allocated budgets as a result of the economic recession (Nicholas et al., 2010; Lindqvist, 2011).

Second, the maintenance of the above-mentioned IT-infrastructure for access and distribution of digital information resources is not only expensive; it requires an integrated information strategy as well. Library management for a long time considered information technology and information services to be two separate resources, and did not recognize that convergence could create efficiencies or synergy (Massis, 2011). Moreover, the academic libraries’ organizational culture in general is found to be averse to technological innovation (Jantz, 2012). As a result a robust IT infrastructure may not be recognized as a required resource for online access and distribution.

(9)

In order to acquire especially the financial resources necessary to retro-digitize their backlist, academic libraries defined partnerships with private parties such as publishing entities as a strategic goal.

3.2 ACADEMIC LIBRARIES’ KEY PERFORMANCE INDICATORS

As in the private sector, performance measurement in public organizations, with the aim to transparently track and assess, will become the more and more important (Jantz, 2012). For academic libraries, as public organizations, developments in public administration with regard to performance measurement are relevant. The so-called New Public Management (NPM) regime, originating in the 1970’s, resulted in guidelines and instruments such as assessment by results, quality programmes, and performance contracts, which, with some delay, were introduced in the cultural sector as governing instruments as well (Ulldemolins & Arostegui, 2013).

As performance measurement, however, is much more extensively researched in the private sector than in the public sector, performance indicators usually focused largely on profitability aspects, such as economic value added, return on investment or profit margins (Turbide & Laurin, 2009). Kaplan & Norton’s balanced scorecard concept (1992) widened the original profit orientation to aspects such as quality, intellectual capital and human resource accounting (Ceynowa, 2000; Gstraunthaler & Piber, 2012). The main objective of the balanced scorecard is to align an organization’s strategic objectives to a selected practical set of performance indicators to critically evaluate the organization’s operational performance (Ceynowa, 2000; Self, 2003). Three conclusions regarding performance measurement in public organizations are supported in the literature. First, measurement aspects include both financial and non-financial indicators; second, performance measurement systems should address both short-term and long-short-term indicators; and finally, performance measurement indicators should be linked to the firm’s mission and its strategy to implement this mission (Turbide & Laurin, 2009).

(10)

From the second half of 1990’s onwards, the political and administrative perspectives on the accountability of academic libraries required more transparency (Hufford, 2013). At this time, therefore, the original balanced scorecard concept was translated to the specific context of academic libraries, with the aim to communicate these libraries’ performance activities transparently to both internal and external decision-making bodies (Ceynowa, 2000). This specific version of the balanced scorecard contains about twenty performance indicators, categorized as financial and human resources; supply of library products, facilities and services; efficiency of internal processes; and use of library products, facilities and services (Voorbij, 2009).

Specific indicators within the scope of this thesis’ research, i.e. the creation of online access and re-use of academic libraries’ backlist information resources, include availability quota; proportion of logins to electronic services from outside the library to total number of logins; cost for each case of use; and the relation of staff cost for electronic services to processing and the provision of electronic services (Ceynowa, 2000). Self (2003) explicitly includes two

complementary performance indicators relevant to the present research, viz. finding money outside the library; and the number of digital materials that were scanned and added to the digital repository according to rigorous standards.

The first proposition therefore is the following:

P1: Libraries aim at maximal online access and re-use of information resources that result from retro-digitization partnerships

3.3 FUNDING DIVERSIFICATION

As described above, academic libraries clearly express their concerns about limited financial resources (Nicholas et al., 2010; Kaiser, 2012; Savenije et al. 2012; Leonardo, 2012), the more so since developments, such as the above mentioned changing user expectations regarding

(11)

access and availability of content, require a shift in budget allocation to digital collection building (Hodges et al., 2010). Also, government cutbacks are expected because of the economic recession (Nicholas et al., 2010; Lindqvist, 2011).

To address this, libraries rely in the first place on strengthening existing public partnerships, and cooperation in national and international library consortia (Nfila & Darko-Ampem, 2002; Nicholas et al., 2010; Giesecke, 2012). Although this strategic cooperation for example results in a stronger bargaining position, it seems that developing financial resources is not a topic on the consortia’s agendas even though this is found to be a necessity in a financial sustainability strategy for cultural institutions as diversified revenue sources increase organizational

sustainability (Carroll & Slater, 2009).

Moreover, not all public organizations are able to develop a diversified financial strategy. A lack of a clearly defined organizational mission and vision may result in an inability to develop an integrated organizational and financial strategy (Ceynowa, 2000; Stanziola, 2012).

Regarding their funding, museums, libraries and archives since the 20th_{century have}

depended largely on government allocations, and still tend to do so (Lindqvist, 2011). A tradition of “ring-fenced government support” (Stanziola, 2011, p. 116) has biased the outlook on funding decisions in the public sector into a preference for cost reduction strategies, political lobbying, cooperation among cultural institutions, and downsizing (Palmer, 1991).

Furthermore, libraries compared to museums and archives show a strong sense of entitlement to government funding, and at the same time a low reliance on alternative funding (Stanziola, 2012).

Since the most recent economic recession that started in 2008, however, governments have stimulated (academic) libraries to enter into public private partnerships to find private funding (Ryden, 2013). As a consequence, the more and more initiatives are visible to seek

(12)

Considering the libraries’ concerns about their financial position together with governments’ stimulation to seek alternative funding, the second proposition is the following:

P2: Libraries for retro-digitization projects actively seek private funding as part of their financial strategy.

3.4 REQUIRED CAPABILITIES IN PARTNERSHIPS

Public-private partnership, in the definition of Brinkerhoff & Brinkerhoff (2011, p. 3), implies “a cross-sectoral relationship where the actors involved bring both commitment and competence to the table, thereby creating the classic synergy (the whole being more than the sum of the parts)”. Partnerships between a public institution such as an academic library, and a private partner such as a publishing entity, do not go without saying however. Private investors usually consider the public sector relatively unattractive because of the required highly specialized competences, high fixed costs, and complex institutional frameworks (Ferlie et al., 1996, in Dubini et al., 2012). Public organizations often regard cooperation with a private partner antagonistically, because of fears and assumptions about opportunistic behaviour on the private partner’s part (Dubini et al., 2012). Moreover, the partnership may be negatively influenced by the public partner’s relative ignorance regarding how to negotiate, to coordinate, or to audit its private partner (Samii et al., 2002). The public partner also often is not used to clearly specify its expectations regarding quality or project outcomes.

As a result of the on-going technological developments, this specific kind of partnerships emerges in a changing landscape, moreover, which implies that, in order to create sustainable advantage, both parties should consider their their pre-partnering dynamic competencies, such as assessing similarities in organization culture or consistency of vision, and capturing

(13)

Other identified key competences required to building a sustainable partnership include mutuality, i.e. a shared goal, shared operational control, and shared responsibilities; clear role definitions of both partners; timing; resource allocation criteria; commitment symmetry;

communication; and active managerial involvement (Brinkerhoff & Brinkerhoff, 2011; Dubini et al., 2012; Samii et al., 2002).

This leads to the following, third, proposition:

P3: Libraries align the partnership set-up to project outcome and performance.

3.5 ADVANTAGES AND DISADVANTAGES REGARDING

RETRO-DIGITIZATION PARTNERSHIPS

As described above, retro-digitization from the academic libraries’ perspective is regarded as essential to create online access to backlist information resources and meet the library user communities’ demands (Hodges et al., 2010). Creating online content not only serves research and education, it also grants access to remote visitors – thus widening the opportunities for public outreach (Kenyon & Hudson, 2004). At the same time, high quality digitization is a means to preserve fragile, deteriorating or rare collections (Power, 1997). Moreover, retro-digitization may stimulate new kinds of re-use (Henry & Smith, 2010).

Emphasized advantages of retro-digitization partnerships from the perspective of the academic library partner are the online access to intellectual content otherwise not digitized, the ease for researchers to find and combine historical documents, and especially regarding Google the volume and speed (Cohen, 2010; Kaiser, 2012). As described above, financial benefits play an important role in the decision to enter into a partnership (Kaiser, 2012; Leonardo, 2012;

Savenije & Beunen, 2012).

Regarding disadvantages of retro-digitization partnerships, especially the literature of participating in the Google Book Search Programme, and the Google Book Search Programme Settlement that was the result of several lawsuits, is extensive.

(14)

Topics discussed are for instance erosion of copyright, orphaned works, and exclusivity terms (Band, 2009; Hanratty, 2005; Hausman & Sidak, 2009; Muller-Langer et al., 2011, to name just a few). Noteworthy with regard to this thesis research is that Google changed its contract to the benefit of the libraries that joined the programme from 2008 onwards. Interestingly, in the literature hardly any discussion was found on other partnerships and contracts, even though the contracts in several aspects may be found to be unfavourable compared to the adapted Google contract. Contracts and conditions are further discussed below.

3.6 GOALS TO ACHIEVE IN RETRO-DIGITIZATION PARTNERSHIPS

A set of common topics the potential partners should negotiate when entering a partnership for retro-digitization in the literature is clearly defined (Johnson, 2007; Pike, 2012; Bell, 2012). Strategic business issues range from ownership and intellectual property rights; exclusivity agreements and terms; maximization of access and re-use for different user communities; the roles of both partners in the project; and control of brand name. Expert library issues include long-term accessibility, quality issues regarding the digital scans and related metadata, and technology standards (Cohen, 2010; Leonardo, 2012).

From the academic libraries’ perspective, goals to achieve in contract negotiations include the non-exclusivity of the partnership (i.e. the library is free to enter a partnership with other parties to digitize the same content); ownership of all digital copies for the library to use for online access purposes free of charge, including access through publicly funded platforms and to research partners; the obligation to both parties to make the content available free of charge; autonomy of the library to select the materials to be digitized; supervision and evaluation of the logistics and digitization processes by the library; and the possibility to terminate the

partnership if the results do not meet the library’s expectations (Kaiser, 2012). As exclusivity, term, ownership, and property rights are particularly relevant to the libraries’ opportunities to

(15)

create online access to existing and new user communities, these aspects are further investigated.

3.6.1 Exclusivity Agreements and Term

To protect investments and to prevent competitors to easily copy a product, businesses may use exclusivity as a way to prohibit a party to do business with a third party (Segal & Winston, 2000). For a library, on the other hand, exclusivity proves difficult as it implies that its

collections are closed to all other potentially interested parties – which goes against the institution’s basic principles (Bell, 2012). From the perspective of the library, therefore,

exclusivity should be minimized as much as possible in order to prevent restrictions on re-use or to hinder future opportunities to enhance value (Johnson, 2007).

To meet halfway, a possible solution is to include a clause in a contract to grant the private party preferential access for a certain term, as well as a clause to ascertain that an interested third party may digitize the same materials again. This latter clause will be acceptable to the private party, as the market for digitized backlist library materials is highly specialised. The chance that another commercial party will be interested in scanning the same materials for a second time therefore is negligible (Bell 2012).

What is more, in June 2013 the European Union published an amendment to its Directive on

the Re-use of Public Sector Information (PSI), and extended it to heritage resources. Article 31

explicitly states that an exclusivity term in case of retro-digitizing material already in the public domain should not exceed a period of ten years.

The fourth proposition therefore is the following:

P4: Whereas publishers will aim to negotiate the longest possible exclusivity term, academic libraries will tend to negotiate the shortest possible term.

(16)

3.6.2 Ownership, Property Rights and the Public Domain

In order to create profit, owners of property divide or combine resources (Heller, 1999). Essential to property rights are the owner’s exclusive right to use his resources as he sees fit, and his ability to voluntarily share or transfer them (O’Driscoll, 2003). Kinds of ownership vary from open access (cultural commons, or public domain) to full exclusion (anti-commons) (Cooper, 2011; Heller, 1999; Savenije & Beunen, 2012). Owners can exclude others from using their resources, except for resources under open access (Heller, 1999). In this research, especially aspects related to property rights and open access are relevant, as retro-digitization efforts often concern materials that already are in open access.

Intellectual property rights legally are regarded as monopolies of restricted duration (Savenije & Beunen, 2012). After the expiration date, the work falls into the public domain, which means that it is free for all kinds of re-use such as reproduction, adaptation, distribution, and display (Cooper, 2011; Savenije & Beunen, 2012). Defining the public domain, however, proves difficult since the introduction of the Internet, digitization and the resulting online access to, for example, art through digital reproductions (Towse, 2008; Cooper, 2011). Regarding digitization of materials in the public domain, several aspects concerning property rights are a matter of debate. The first aspect is whether the creation of digital scans leads to new copyrights for either the content owner who ordered the digitization or the company who created the digital scans (Savenije & Beunen, 2012). The position of libraries, and other cultural institutions, in this respect is quite unique as they are content owner and at the same time distributors of digital content (Kenyon & Hudson, 2004). Private companies that created the digital copy, however, tend to claim copyright protection over these copies. Requiring payment for access or forbidding re-use results in restricted access to works already in the public domain (Cooper, 2011), which, as mentioned above, goes against libraries’ principle that materials in the public domain should be freely accessible to all parties (Bell, 2012).

(17)

Opinions vary, moreover, whether the digitization process adds creativity to the scan through for instance the creation of additional metadata for access purposes, which might imply new property rights (Savenije & Beunen, 2012).

Regarding property rights, the above mentioned PSI-directive states that materials of which (academic) libraries hold these rights, i.e. materials in the public domain, should be made available for all forms of re-use, including commercial purposes.

A second aspect concerns database rights related to the investments regarding creation, presentation and verification of all the content in a database together with its websites for access (Savenije & Beunen, 2012). Database rights protect the investor from re-use without permission of substantial parts of the content only.

This leads to the fifth and final proposition:

P5: Academic libraries claim property rights of the created digital files as well as its accompanying metadata, and database rights.

3.7 RESEARCH QUESTION

As described above, academic libraries enter into partnerships with publishing entities for retro-digitization purposes for two reasons. First, they partner as a part of their funding

digitization strategy. Second, creating online access to their collections is a necessity in times that users the more and more expect 24/7 unrestricted access. In order to optimally perform their tasks, libraries specifically need to negotiate exclusivity terms and property rights while entering into a partnership with a publishing entity, as these are aspects that will to a large extend determine the library’s user communities’ level of (free) access to this content, and the libraries’ options to develop new services to existing or new users.

(18)

The research question therefore is phrased as follows:

How can academic libraries that, for financial reasons, enter into retro-digitization partnerships with publishing entities to create online access to, and to stimulate re-use of, their backlist information resources, settle on the contractual aspects exclusivity, term of the cooperation, ownership, and property rights rights in a way to enhance their competitive position?

This research question arose from a practioner’s interest in the way in which libraries and publishers cooperate to digitize backlist information resources for access and re-use purposes. How libraries are managing this kind of partnership to their advantage, and how they use the project results with regard to their competitive position, are topics not yet widely researched, neither in the field of library sciences nor in business studies.

(19)

4. RESEARCH DESIGN AND METHOD

This thesis is designed as explorative research, which in the definition of Saunders & Lewis (2012) is “research that aims to seek new insights, ask new questions and to assess topics in a new light” (p. 110). The method chosen to conduct the research is a multi-case study, as the research question is aimed at a contemporary topic within a real life context (Saunders & Lewis, 2010). Given that empirical research from the perspective of a public partner is as yet not widely adopted in business studies, this research is approached inductively. As transparency and replication are of utmost importance in especially qualitative research, this method section serves the purpose of the research protocol (Gibbert & Ruigrok, 2010).

4.1 STRENGHTS AND LIMITATIONS OF CASE STUDIES

Generally speaking, the strength of case studies is the possibility to collect rich data as the basis for a qualitative representation of evidence (Eisenhardt, 2007). Considering causal forces, case study data provide a more persuasive argument (Siggelkow, 2007). From a practioner’s point of view, it is also the most rewarding way to conduct research, as it offers the opportunity to contribute largely to organizational and managerial practices (Gephart, 2004). A multi-case, the method of this research, will result in more accurate, generalizable, and better-grounded research propositions based on rich empirical evidence (Eisenhardt, 2007). As a multi-case resembles multiple experiments, analytic generalization will result in the comparison between previously developed theory and the empirical results of the cases (Yin, 1994).

Limitations to case study research are quite clear, however. Strict norms regarding rigour, validity, generalizability and reliability of the research must be applied to facilitate

(20)

replication and address possible bias from quantitatively oriented researchers towards objectivity of the research (Gibbert & Ruigrok, 2010; Siggelkow, 2007). Possible flaws of case study research include the lack of balance between theory and collected data (Pratt, 2009), and sample-size (Siggelkow, 2007).

4.2 LIMITATIONS OF THE PRESENT RESEARCH

Limitations to the present research are twofold. First, the perspective of the private parties involved in the three examined cases is partly represented, as only publisher Brill

accepted the invitation to participate (see below in paragraph 4.4). Second, several respondents referred to confidentiality clauses in their respective contracts, with the result that contract copies, project plans or internal reports generally were not obtained.

Financial details and exact data on library performance, therefore, were not collected. This was addressed as far as possible by using aggregated data from policy plans and annual reports available online.

4.3 RESEARCH SETTING

The results of the theoretical research were applied to three cases. The first case is the partnership of the Royal Dutch Library (Koninklijke Bibliotheek, KB) with Google that started in 2010 and will continue until 2015, as part of the Google Book Search

Programme. At the end of this project, about 160,000 19th_{century Dutch books will be}

accessible online. The second case is the partnership between KB and publishing entity ProQuest, in which 30,000 Dutch books from the 16th_{and 17}th_{centuries are digitized. This}

project also started in 2010. The third case is the partnership between the library of Leiden University (UBL) and publisher Brill. They recently started their third retro-digitization project, and are negotiating a fourth. The results of the first two projects, about 150,000 digitized manuscript pages, are part of the Brill online resources portfolio since 2011 and

(21)

2012 respectively.

The cases were selected because they represent different angles to success or failure of cooperation between an academic library and a private partner, i.e. the selection is information-oriented (Flyvbjerg 2006). The specific respondents were chosen because of their relevant practical knowledge and experience on a strategic organizational level with regard to the research question.

The cases are complementary in the sense that the cooperation between UBL and Brill is an example of a long lasting cooperation that therefore seemingly is advantageous to both partners. The partnership of KB and ProQuest is probably a one-time only effort of limited duration. The case of the Google Book Search Programme represents the biggest retro-digitization effort to date, with Google aiming to digitize 15 to 20 million titles worldwide. Together, the three cases offer a broad perspective on the cooperation between public and private “content providers” in the area of retro-digitization.

4.4 DATA COLLECTION PROCEDURE

Access to the respondents was arranged through personal recommendations and introductions, and direct e-mails to respondents within the researcher’s network. Getting access to the private partner strictly speaking was not included in the research, as the focus is on the perspective of the public partners. Nevertheless, in order to triangulate the results, the perspective of the private partner is of interest. Access to publisher Brill was obtained; Google and ProQuest were both approached by email, but both declined the invitation to participate (emails in the researcher’s possession).

The primary data were collected through semi-structured interviews with project

management and senior-management of the above-mentioned academic libraries, and publishers. These interviews all lasted approximately one hour. With the aims of

(22)

interviews were all audio-recorded, and transcribed verbatim afterwards. In order to triangulate the interview data more rigorously, in each organization at least two respondents were interviewed.

Access to secondary data from the institutions themselves or contracts between the partners proved more difficult. Aggregated data about the project results and access figures were therefore obtained through online policy plans, annual reports, and websites.

4.5 DATA ANALYSIS PROCEDURE

The semi-structured interviews were prepared in such a way that the transcriptions could be deductively categorized into seven key issues. As a first step in the analyzing process, relevant parts in the transcripts were axially coded (Mortelmans, 2013) into partner choice; negotiation; cooperation; exclusivity and exploitation; ownership; library performance; and financial considerations. Different colors were used to identify the categories in the

transcripts. The unit of categorization was complete sentences. All text units were

numbered uniquely, consisting of a number to identify the respondent and a serial number. Next, the coded text units were converted to an Excel data matrix. For each of the seven initial categories a separate worksheet was created. On the horizontal axis, every

worksheet contained three subgroups, i.e. library project management, library senior management, and publisher. The unit of categorization remained complete sentences. In order to retain the connection between the transcripts and the data matrix, all text units contained the unique reference numbers. Processing all interview data as described above thus resulted in a data matrix that represents selectively coded data (Mortelmans, 2013), and as such facilitated the process of pattern and interrelation detection. In the final step, the most representative quotes were paraphrased in English, and assembled in illustrative tables to be presented in the Findings section (chapter 5).

(23)

5. FINDINGS

In this section the collected interview data are analysed, and compared to the five propositions presented in the Literary Review (chapter 3). The reference numbers in the text below refer to the paraphrased quotes from the interviews. In order to facilitate reading this section, the quotes themselves are assembled in accompanying tables.

The primary data from the interviews are complemented with references to secondary data sources, where relevant.

5.1 LIBRARY PERFORMANCE INDICATORS REGARDING ONLINE

ACCESS AND RE-USE

In the articles on the development and implementation of performance indicators for academic libraries quite detailed information was found on type and usage of indicators with regard to online access, retro-digitization and funding (Ceynowa 2000; Self, 2003; Voorbij, 2009). As the development of performance indicators for this sector started in the 1990’s the expectation was, therefore, that libraries in the meantime would have implemented them, and that sufficient information would be found to draw sustainable conclusions with regard to the effects of retro-digitization on online usage. Surprisingly little information is available, however.

When asked about the results of the digitization projects, and its contribution to online visits as performance indicators, the library respondents, including senior management, answered that working with key performance indicators is quite new to them (3:51, 3:52, 7:46). Remarkable, too, is the absence of performance indicators related to the strategic priorities fund raising and partnering, explicitly described in the libraries’ policy plans.

(24)

Table 1: Performance indicators towards retro-digitization effects on online usage

Subject Exemplary quotations

Use of performance indicators Only this year we really started working with proper indicators [3:51] It proves very difficult to get information on online usage [3:52] How to define access, what is access? [4:32]

Research shows that usage growths once materials are digitized. But what is usage exactly; you can't measure it [7:46]

Regarding the KB, some more information on performance indicators is available in their 2012 Annual Report. The KB defined key performance indicators in 2008, which will be evaluated in 2013 (KB Annual Report 2012, p. 17). Their Type 1 indicators relate to performance regarding the strategic priorities speeding up retro-digitization, and creating access to users. Indicators are for instance online visits to the website, and numbers of digitized collections in publicly funded programmes. Unfortunately, no indicators are included regarding the retro-digitization projects in partnerships. This may be explained by the fact that the indicators were validated in 2008, when this specific type of projects had not yet started. The financial section of the Annual Report does not contain any information about these project results either.

On the basis of the indicator “Online visits to the website” the only viable conclusion is that the number of online visits increased from 3,8 million in 2005 to 5 million in 2012. It is not possible to connect these figures to the figures on digital materials, and draw any conclusions, because these are represented in different units (books, collections, e-journals or e-articles).

The UBL Annual Report 2012 is not available. Their 2011 Annual Report does not contain any specific performance indicators at all. Based on the little information that is available,

(25)

5.2 LIBRARIES’ STRATEGY REGARDING RETRO-DIGITIZATION

PARTNERSHIPS

As described in the literary review, libraries recognize the need to create online access to their collections to facilitate their user communities, but generally lack the means to do so (Kaiser, 2012; Savenije et al. 2012; Leonardo, 2012). Speed and volume, furthermore, are specific reasons for libraries to engage in the Google Programme (Cohen, 2010).

This is reflected in the libraries’ policy plans and long term strategy documents. The KB formulated public-private partnerships for speeding up its retro-digitization programmes as a priority (Long Term Strategy Document, p. 3). UB Leiden formulated fund raising activities as a strategic goal (UB Leiden Policy Plan, p. 14). In the UKB Policy Plan, decreasing financial means, retro-digitization and partnering with commercial parties are explicitly linked (p. 7). All respondents, including the publishers, agreed on libraries needing funds to digitize as a reason to partner (1:6, 1:8, 2:27, 3:10, 4:26, 5:14, 6:62, 7:8, 8:2). Creating online access was less often explicitly mentioned (3:12, 5:10, 7:1). With the exception of R7 (7:2), who is indeed working at a Google library, speed and volume were not mentioned.

More importantly, it is found that libraries in general are not proactively approaching private partners as part of their funding diversification strategy (2:1, 3:1, 4:6, 4:18, 5:1, 6:69, 7:26). This means that the second proposition proves to be incorrect.

It was also found that, despite their concerns regarding partner dependency (Dubini et al, 2012), libraries thus far created online access to their retro-digitized information resources by means of the partner’s infrastructure (2:38, 4:63, 4:70, 7:24). This confirms that libraries thus far do not recognize an access and distribution infrastructure as a strategic resource (Massis, 2011; Jantz, 2012). It may also imply that the libraries did not clearly define their ingest and access routines as part of the projects.

(26)

Table 2: Libraries’ strategy towards retro-digitization partnerships

Insufficient financial resources More importantly, we don’t have a budget for mass digitization [1:6] In all academic libraries you will find there is little or no budget for digitization [1:8]

So we offer them funding, and say: now we want to digitize this [2:37 The next ten years we won’t digitize them ourselves. We lack the money and the means [3:10]

We are looking for, you know, who can help us bear the cost of digitization? [4:26]

We don’t have any money for mass digitization. There is no funding to do that [5:14]

We feel the cost for digital, because it is not included in our budget [6:62]

We were looking for opportunities to digitize, and not having any money ourselves we were forced to look elsewhere [7:8]

They do not have the means to fund the digitization themselves. So we do that for them [8:2]

Creating online access It is a win situation because we create digital access to collections in the public domain more or less for free [3:12]

We are very clear in our mind why we are doing it, it is not about generating income. It is about creating digital assets [5:10]

In our policy plan 2010-2013 we stated that our aim is to have digitized our entire collection by 2025 [7:1]

Volume and speed And I wanted to be more specific, so I said the aim is to digitize 10%, about 70 million pages, by 2013 [7:2]

At the end of the project, Google will have scanned about 30-40 million pages for us [7:17]

Partner selection I think publishers start most of the initiatives, libraries are not as entrepreneurial [2:1]

I believe the publisher approached us [3:1]

With both these partners it is not so much that we chose, but that we are chosen [4:6]

It depends on who the partner is [5:1]

Well, we were not so entrepreneurial ourselves [6:69] They approached libraries, including us [7:26]

Well, both projects were initiated because we didn’t have any money and a third party presented itself [7:36]

Infra (access; distribution) We provided them with copies of our scans on a hard disk, but that’s parked on a shelf and not used [2:38]

We don’t use our own copies, but up till now provide access through Google [4:63]

(27)

Table 2 ctd: Libraries’ strategy towards retro-digitization partnerships

Infra (access; distribution) We show them by means of a Google API, because as yet we have to develop software for ingest as well as access. But the user doesn’t notice that [7:24]

5.3 MANAGING THE PARTNERSHIP

As described in the literary review, partnerships in the cultural sector are often regarded as problematic from both the perspectives of the private and the public partner (Dubini, 2012). Possible drawbacks include a lack of negotiating skills, and an inability to define project

specifications on the part of the public partner (Samii, 2002). Recommendations to ascertain a successful partnership, next to mutuality, include clear selection and assessment routines, and clearly defined roles, goals, and results prior to the start of a project (Brinkerhoff & Brinkerhoff, 2011; Dubini, 2012; Eisenhardt, 2000; Samii, 2002).

The respondents’ answers, despite a wide variation in partner type, show a great deal of similarity in explaining why they consider their particular partnerships a success. Even though they generally did not actively approach their partner themselves, all library respondents could clearly explain why they found the offer to partner attractive (1:32, 1:33, 3:2, 3:3, 6:16, 6:27, 7:6). This varied from an intimate knowledge of the particular publisher resulting from prior co-activities, to having a shared goal in spite of the wide gap between the two organizations. In all three cases, the parties were confident about the outcomes of the negotiating phase. Results included for instance royalties during the contract term, or extra budgets for restoration (3:11, 3:16, 3:17, 3:18, 7:9). The results with regard to exclusivity, term, and property rights, as the most relevant aspects determining online access and re-use of the digitized information resources, are described separately in paragraphs 5.3 and 5.4 below.

At the same time, several respondents acknowledged that private parties might have taken advantage, as public parties are known to be inexperienced negotiators (3:23, 6:49). With regard to project management, most of the respondents considered the projects

(28)

themselves to be “no rocket science” (5:30), in spite of volume and acceleration of speed. As a result, usually no formal project management was established. Instead, once the formal work procedures were decided on, in all three cases the private partner coordinated the digitization process quite independently (2:14, 2:27, 2:92, 2:96, 3:28, 3:30, 3:38, 4:12, 8:12, 8:21). Although it is too early to determine if all projects will be successful indeed, as two out of the three cases examined are not yet finished, it seems safe to conclude that thus far there are no signs that might indicate failure. KB and Google are on schedule, as in December 2013

100,000 of the targeted 160,000 books are online (7:24). The cooperation between Brill and UB Leiden may safely be deemed successful, as their first two projects in the meantime are finished, the third started last October, and the fourth is being negotiated on (1:30). Even though no exact figures are available, as both parties referred to confidentiality clauses in their contract, it is clear that the first two projects are successful from a commercial point of view, as a publisher respondent admitted to have reached his break even point within two years (2:97). From the perspective of the library partner, the partnership is successful as they decided to build their own digitization facility, which will be largely exploited in the third and fourth projects (3:46).

Table 3: Managing the partnership

Partner choice We already had contacts with this publisher, and they have an established reputation [1:32]

You can say we have a tradition [1:33]

It is the continuation of a long existing cooperation. We have always needed each other [3:2]

We trust each other [3:3]

The relationship is extremely skewed, Google being one of the largest multinationals worldwide [6:16]

But in spite of this skewed relationship, both parties get something very valuable to themselves but of little interest to the other party [6:27]

(29)

Table 3 ctd: Managing the partnership

Partner choice We met, and we had a click. The personal aspect is very important [7:6] Negotiation So we wanted some kind of contract in which the publisher paid for

digitization, that they could sell the product, and that we got the content back for free access in the public domain within a term in which we would never be able to digitize and offer online access ourselves [3:11] We negotiated money for restoration of our books, and we got it as extra goodwill [3:16]

And we got royalties for the term of the contract [3:17]

We’ve been bickering about the contract for quite some time. I think we came out well, because we already had an intimate relationship to start with [3:18]

Well, ofcourse they will try to take advantage of the lack of commercial insight typical of our kind of institutions. We aren’t skilled negotiators. We don’t bring lawyers to the negotiating table, as we probably should [3:23]

I did the negotiations and sometimes I asked for legal advice. But we don’t employ a specialised lawyer, and the other party is a huge American enterprise that employs a whole department of lawyers [6:49] The contract negotiations were quite easy, because I had formulated in advance the essentials I wanted to be part of the contract [7:9]

Project management We have very short lines, I’ve known them since twenty years. So I just phone and talk to them [2:14]

Well, that all was quite organic. Yes, you can say it went pretty smoothly [2:27]

You are partners. You need to take each other’s wishes into account. It's no use to be very rigid [2:92]

It is a well-understood self-interest on both sides. And if you can organize it in a way that benefits both parties it’s fine [2:96]

You can say the publisher managed the project [3:28]. This was decided on because I considered it to be a relatively easy project [3:30]

We didn’t evaluate, because I thought this was not a difficult project [3:38]

We really looked into it beforehand, what is feasible and what not [4:12] It is all about trust. And you can do that because you have a working relationship [8:12]

(30)

Table 3 ctd: Managing the partnership

Project management The lines are very short, and it works well. If something comes up, we can switch very easily, and arrange something else [8:21]

Success of the projects We decided on three projects, we signed three contracts. And two of them have been executed [1:30]

We’ve reached the break-even point. For this kind of projects you estimate your break even after a couple of years [2:97]

Now that we have our own digitization equipment, financially it is even more attractive to us. The project helped us prepare our own

organization, you know, project administration, training people, working with deadlines [3:46]

This means that the third proposition is only partly correct. Project management is clearly defined as the role of the private party. The project management structure itself in all three cases is quite loosely arranged, however, without (quality) checks or evaluations. In spite of this, all respondents declare their projects to be successful. This may be explained by the fact that all parties, publishers and libraries alike, prior to these specific projects had already participated in other retro-digitization projects. As learning effects play an important role in developing organizational skills (Zollo & Winter, 2002), these prior experiences may explain why respondents treat their projects as business as usual. In the case of UBL and Brill, inter-firm trust may also be an explanation (Das & Teng, 2000), as both parties have known each other for quite some time.

5.4 EXCLUSIVITY, TERM AND THE LIBRARY BUSINESS MODEL

From the libraries’ perspective creating online access to facilitate research is the primary aim to digitize (Hodges et al., 2010; Kaiser, 2012; Kenyon & Hudson, 2004). They cooperate with private partners to finance the digitization process. In order to allow the investing publishing entity to make a return on investment, an exclusivity agreement generally is agreed on for a

(31)

certain period. Of the three models examined, the Google model is the only one that is non-exclusive.

The terms and conditions of the three researched models differ considerably, however. The terms of the exclusivity vary from 10 (Brill) to 15 years (ProQuest), during which period libraries are allowed to provide free online access to their academic user communities. This usually is arranged through their online catalogue on the library website (1:38, 3:14, 3:41, 4:14, 6:42). The Google model is also the only model that allows free access to a wider audience from the start (2:91, 4:14, 5:6), but for non-commercial purposes only. This means that the content, next to access through the libraries’ website, can be made available worldwide through portals such as Europeana. Interestingly, the KB with ProQuest negotiated access to the Dutch general public instead of their existing academic user community only, which in practice means they will have to find a way to restrict access to IP-addresses in the .nl domain (6:42, 7:28). All other parties, including the Dutch academic libraries, will have to pay for access to the digital content, through outright purchase (perpetual access) or yearly licenses (8:24, 8:25). The ProQuest scans are also offered to libraries as a part of the so-called big deals (7:29). Commercial exploitation by the library in all three models is not an option as long as the exclusivity term is valid (1:37, 2:84, 3:19, 4:37, 6:41, 7:34).

Table 4: Exclusivity and the Library Business Model

Exclusivity We have an exclusive licence; otherwise it won’t work for us. If we allow them [the library] to sell the scans to a commercial party, we ruin our own market [2:53}

I mean the first huge difference is that the Google content is all free. And with all the other partners it sits behind a pay wall, so it’s a massive difference between the Google model or the ProQuest model, or the who else eh the Brill model or Sage [5:6]

(32)

Table 4 ctd: Exclusivity and the Library Business Model

Exclusivity I believe we have a great contract compared to other ProQuest libraries. We limited the project term to 2 years. If they don’t finish the project in 2 years, the exclusivity term starts anyway. The exclusivity term starts the moment they put content online or after 2 project years, and not after 7 years as they first proposed. So the exclusivity term has already started, because they are not yet finished. It will take them probably 4 to 5 years [6:49]

Well, personally I have some reservations. Especially this

non-exclusivity part, well… you put part of your collection behind bars for the next fifteen years, and you can’t exploit it yourself. You are completely dependent of the supplier. I mean, it sounds great free access in the Netherlands but in a way it doesn’t feel good [7:38, on ProQuest] Public exploitation We have created online access to members of our academic

community. They have free access [1:38]

Library members have free access. So everybody who holds a library pass has free access, and I am quite certain everybody can buy a library pass [3:41]

We can offer free access to the Dutch public; the rest of the world has to pay. We can only make it freely available in fifteen years time [4:14, on ProQuest]

Well, we discussed who could get free access, the library community or the Dutch general public. I said it is a Dutch collection, so I want all Dutch citizens to have free access [6:42, on ProQuest]

Commercial exploitation What I couldn’t allow was the library to give the scans away to who knows whom, and put them in open access. That is impossible [2.91] Google allows us to share all content with other libraries, no problem. What they don’t allow is commercial services for the next fifteen years on the scans they invested in. They don’t want other search engines indexing the same corpus [4:3]

We find commercial exploitation quite difficult as yet, because it can create some kind of precedent. If we sell it for a certain price and another commercial party is interested, do we always have to ask money for the datasets? And the other way round, the open access directive will restrict us in what we can do. The tendency is to allow access to all parties at reasonable cost, including commercial re-use [4:29]

If it is commercial, it is a whole differing thing. Than we would put them together with the publisher and say here, you sort it out, or not [5:25]

The forth proposition proves to partly correct only, therefore, as only the Google contract is non-exclusive, and the two other contracts are. As Google changed its terms after several lawsuits, and uses the same contract in its worldwide programme, it is safe to conclude that this non-exclusivity is not the result of the negotiations in this specific case. This is confirmed

(33)

by R7, who was responsible for the negotiations with Google (7:14).

The ProQuest contract, moreover, has a term of fifteen years, five years longer than the PSI-recommendation.

What is more, libraries are found to see little commercial value in their (digitized) collections, which is expressed in a surprise or disbelief in positive project results, and references to high financial risks on the part of the private party (3:38, 5:31, 5:32, 6:32, 6:33, 6:36). This is contrary to the remarks of the two publisher respondents, who are quite confident about the value of the collections they digitize, market potential and profit (2:5, 2:48, 2:69, 2:75, 8:23, 8:28).

Although exact figures on the profitability of the first two UBL-Brill projects are unavailable because both parties referred to confidentiality clauses in their contract, it is possible to deduct a rough estimate on the basis of the Brill Online Resources Catalogue combined with interview data. The results of the two projects are offered on sale as part of the Middle Eastern

Manuscripts Online portfolio (Brill Online Resource Catalogue, p. 97). The first set, consisting of 110,000 pages, is on sale for € 20,000; the second set of 46,000 pages is on sale for € 8,840. Both are offered as outright purchase only. Respondent 2 mentioned he expected to sell both sets to approximately 30 institutions worldwide (2:69). A rough estimate of the expected turnover, therefore, is about € 865,000. Respondent 8 mentioned that the scanning price per page is € 0,40 (8:19). Scanning costs therefore estimate to € 62,400. Allowing for further upfront costs for quality checks and creation of metadata, as well as fixed costs for database and infrastructure maintenance, marketing and sales etc., it is nevertheless clear why, from the publisher’s point of view, both projects are worthwhile.

(34)

Table 5: Experiences in current projects

Recognizing value But the publisher said to me about the sales of the first product: it did extremely well. And he also showed a little surprise [3:38]

I mean we are realistic about the fact that they need to realise a decent return. So there has to be a meeting in the middle [5:20]

Sometimes I can’t believe that a publisher is making any money at all in a project [5:31]

There is no commercial value in 19th century literature. I wish all parties who want to create commercial services the best of luck [6:32]

And the investment is high [6:36] Publisher’s perspective Because there is a market [2:05]

It’s an economic consideration. If we don’t think we can sell it, we don’t do it. It’s as simple as that [2:46]

We sell it because it’s this library’s collection. We select well [2:75] So it is very interesting to us, because we get something that in the academic world is recognized as one of the most important manuscript collections [8:1]

Most of our potential buyers are from the States. The world is our market [8:23]

We operate in a niche market. It depends on the product where our main market is [8:28]

It is possible to develop a transcription product, and crowd sourcing [8:29]

Both publisher respondents also were clear about developing new business with digitized content. They are – within two years after the start of their first retro-digitization project – exploring new business opportunities such as developing OCR software or crowdsourcing projects (8:29). Libraries seem to be more hesitant when they are approached with similar requests by either commercial or public parties, or find that the contract is not clear enough about new ways of exploitation such as software development, or data or text mining (1:27, 4:58, 4:68, 6:5, 7:52). If they do see options for exploitation, they emphasize free online access to the digital content, and sharing publicly owned materials for free. Furthermore, they refer to the public nature of their organization, the EU PSI Directive, and their public financier (1:15, 1:23, 3:54, 4:30, 5:8, 6:9, 7:54, 7:59). This may confirm Stanziola’s (2012) findings about

(35)

libraries’ outlook on financial diversification strategies, and their tendency to rely on public funding.

Table 6: Library Business Options

New business options Actually, we only just started to investigate [1:27]

Well, the contracts are a point really. We are the more and more aiming at re-use of materials, and how to agree on it as clearly as possible [4:58]

Reaction to options We see now that the clearer you define re-use in advance, the less haziness there will be on, well we have a contract but what are we allowed to do? [4:68]

Cooperation in a European consortium will work out, but as soon as a commercial party joins the development or brings his own software, things get complicated [6:5]

Commercial parties are willing to pay money for the data. And they need to make a profit, fine. But, well, we have to share everything for free, we believe. That’s fairly principle [7:54]

It is important to create open access [1:15]

As a library, the Ministry of Education, Culture and Sciences is our financier, but not the C for Culture but the S for Sciences [1:23] You have to take financing models into consideration. The university pays us, and we are not paid to make our collections accessible to for instance the elderly [3:54]

And if we say yes to a commercial party, won’t it imply some kind of government support to those businesses? [4:30]

PSI is the Public Sector Information Directive, which never used to apply to our sector, to libraries and archives. We are now brought in to that directive and the essence of it is transparency [5:8]

It is ours, but we share it with everybody. That’s fine with us [6:9] The Higher and Scientific Education department of the Ministry is our financier [7:59]

5.5 PROPERTY RIGHTS AND OWNERSHIP

In all three cases it was found that ownership of the created digital assets lies with the private party (3:15, 4:4, 4:50, 5:11, 5:12, 6:31, 7:10, 8:3), which is contrary to the recommendations

(36)

found in the literature (Bell, 2012; Johnson, 2009; Kaiser, 2012). However, in all three cases copies of the files are provided to the libraries, although exploitation usually is restricted during the contract term. Possibly, as the debate on ownership of digital assets is not quite settled (Savenije, 2012), libraries consider independence of the private partner on the longer term, which is granted by this copy, more important than executing property rights on materials burdened by exclusivity agreements on the short term (1:16, 1:18, 4:47, 5:33, 7:49). Concerning database rights, one of the libraries is allowed to sell single digital assets to researchers, and has to refer to the publisher for more extensive requests (1:42, 2:40, 2:43). This might indicate that the publisher executes database rights (Savenije, 2012).

Proposition 5 proves to be partly correct only, therefore. Property rights on the created digital assets are not obtained, property rights on metadata are.

Table 7: Property Rights and the Library Business Model

Digital assets In the contract it is stated that the publishing entity owns the original because he financed the project, and that we get a duplicate in the format we require [3:25]

We get a copy from Google [4:4]

We get a copy [from ProQuest] but for the time being we are not allowed to do anything with it [4:50]

Well the Google model is one where you own the copy. And you don’t have the terms and conditions on what you can do with that [5:11] I think it's much better when you own the digital asset. It is in a grey area of the law, because somebody will say you can’t say you own a digital asset [5:12]

They [publishing entity] own the scans [6:31]

Everything that is scanned is made available to us [7:10] They [library] get an electronic copy of everything we scan [8:3] Metadata The library gave us a licence, not an exclusive one, but we are allowed

to publish them together with the scans [2:22]

We agreed on ownership of the metadata, they remain ours. We arranged different aspects; we included them in the contract because it was all new to us. Who owns the metadata, who owns the metadata created in the project [3:24]

(37)

Metadata are free, that is our philosophy. We own them but we share them. All metadata are free except when the publishing entity explicitly prohibits it [6:8]

Database rights We are allowed to sell single scanned pages, but for entire data sets we refer to the publisher [1:42]

Table 7 ctd: Property Rights and the Library Business Model

Private partner dependency … that everything will be in one hand and some fee will be charged. Who can guarantee that that won’t happen? [1:16]

And well, we need to be sure that if you do all those investments, that there will be a permanent infrastructure [1:18]

Sometimes I’m afraid Google will end the project just like that [4:47] I mean you think Google has been around forever, but it is only sixteen years. So we said have you ever done a thing like that before. We are a cautious organization [5:33]

We will always facilitate some sort of basic accessibility, because we don’t want to be dependent on an external party that will or won’t be interested in investing [7:49]

Concerning future possibilities for privately funded retro-digitization projects, and thus the sustainability of this kind of partnerships, respondents mentioned copyrights as a major barrier as it will preclude commercial online exploitation of all materials from the larger part of the 20th century, the bulk of all materials stored in libraries (5:34, 7:39). This means that the number of potentially interested private parties is expected to be low (1:50, 7:40, 7:45).

Table 8: Perceived barriers to future backlist digitization projects

Copyrights What one really really wants is 20th century content, but most of this content we can’t really touch because it is so complicated in terms of clearing rights [5:34]

Well, the interesting period ofcourse is the second half of the previous century, until now. That is material burdened heavily by copyrights, but it is by far the bulk of our collections [7:39]

Competing for access and re-use? : retro-digitization partnerships in Dutch academic libraries.