3 Theory review

The previous chapter described the influences that steer academic publishers toward a new business model in which free online access plays an important part. In order to research the implications, it is important to focus on the most relevant aspects of this new model. In this chapter, the role of Open Access publishing is discussed using themes related to the library and information sciences and the theories in the field of e-commerce.

3.1 Library and information sciences and research impact

Publishing in Open Access is debated widely in the library and information sciences literature, where it is positioned as a solution to limited accessibility to scientific results (Delamothe & Smith, 2004). The term ‘research access/impact problem’ is used to define the negative effects of inaccessible scientific articles through loss of potential research impact (Harnad et al., 2008; Harnad et al., 2004). When an article is not perfectly accessible – although it is available online – its contents do not reach all of its potential readers, causing a lower ‘research impact’. The authors define research impact as ‘the degree to which [the article’s] findings are read, used, applied, built upon, and cited by researchers in their own further research and applications.’ A lower research impact does not only hamper the progression of the scientific field, but also has consequences for the author of the article. Scientific careers are based on the level of impact a researcher has on his/her field, which is commonly measured by the number of citations from a researcher’s work published in a peer reviewed article. If an article is used in further research, the publication containing the results will also contain the citation to the used article. Imperfect access decreases the number of readers and therefore the number of citations.

According to the authors, inaccessibility is caused by the rising prices of journals, forcing libraries to select a limited number of subscriptions. The solution they propose is to make all articles freely available, either directly through Open Access publishing (the ‘golden road’) or by allowing the authors of the article to place a copy in an OAI compliant repository (the ‘green road’). The acronym OAI stands for Open Access Initiative, and describes a protocol designed to enable exchanging metadata and contents between repositories, thus creating a global ‘library’ of freely accessible documents.

3.1.1 Research impact and discovery

In the library and information sciences, much emphasis is placed on citation advantage where articles published in Open Access are generally more cited than articles published in closed access. Discovery and – online – consultation are seen as implicit factors and are generally not discussed. A greater availability due to Open Access removes barriers for the prospective readers. The first barrier to remove is ignorance of the existence of a publication. In other words: before a scientist may consider citing a certain publication, he or she must first be aware of its existence. Publishing in Open Access means that the publication is not only open to human readers, but make the publication also available for search engines. Better availability should therefore lead to more indexing by search engines, which should lead to a higher discovery rate. This leads to Hypothesis 1: The discovery of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

3.1.2 Research impact and usage

After the discovery of a publication, the next step of the prospective reader is to actually get access to the contents of the publication. Here, Open Access publishing may remove another barrier. If a publication is available online in closed access, it can only be accessed if certain conditions are met. In most cases, the publication is made available through a publisher’s system. Access to the document is based on the fact that the reader – or the reader’s institution – gets permission from the publisher, in exchange for a fee. When the reader – or the reader’s institution – does not have such an agreement, the publication is not accessible. Open Access publishing does not use such restrictions, enabling direct access to the document. This leads to Hypothesis 2: The online consultation (e.g. pages read or number of downloads) of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

3.1.3 Research impact and citations

When a scientist has read an academic publication, he or she may decide to use it in his or her own publication, resulting in a citation of the publication. The influence of a certain publication can be measured by the number of citations it received: a higher citation rate means a greater scientific impact. Research indicates that articles available in Open Access have a higher impact than articles available through closed access. This effect has been measured in scientific fields as diverse as philosophy, political science, electrical and electronic engineering and mathematics (Antelman, 2004). Furthermore, articles published directly in Open Access are more immediately recognized and cited than articles published in closed access (Eysenbach, 2006). While there is very little research available on the effects of Open Access publishing of books, the same advantages and difficulties associated with open access articles arise (Bazerman, Blakesley, Palmquist, & Russell, 2008; SPARC, 2009). According to the authors, Open Access books disseminate scholarly work more quickly than print books and appear to be viewed by a larger audience. As a result, these books are cited more widely than comparable print books. This leads to Hypothesis 3: The citation rate of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

As writing a book – the format used most in the humanities and social sciences – will take considerably longer than writing an article, the effects may not be visible within 9 months, the duration of this experiment. Furthermore, the average number of citations varies by different fields of science (Podlubny, 2005). This indicates that differences in citation behaviour may exist between fields of science, which have to be taken into account.

3.2 Transaction costs and information technology

In the field of e-commerce, theories on transactions costs are used to understand the effects of information technology on commercial activities. Publishing in Open Access means that documents are published online, putting the publisher in the realm of e-commerce. After an introduction in transaction costs, the effects on Open Access publishing are discussed in the next paragraphs.

Transaction economics applies to the study of economic organization, with special reference to firms, markets and relational contracting. At the heart of all economic activities lies a transaction. With transactions come transaction costs, which can be described as ‘the economic equivalent of friction in physical systems’ (Williamson, 1998). Economic institutions are created as a response to transaction costs. Transaction costs are divided in coordination costs – the costs before the transaction – and safeguarding costs: the costs after the transaction. Coordination costs are the costs of drafting, negotiating, and safeguarding an agreement. Safeguarding costs occur when transactions drift away from stated contracts, when both parties make efforts to solve these ‘misalignments’, setting up and running dispute resolutions and securing transactional commitments.

The use of information technology (IT) has a profound effect on both the organisation of economic activities and transaction costs. IT has an ‘electronic communication effect’ which increases the amount of information communicated and decreases the costs associated to it (Malone, Yates, & Benjamin, 1987). These effects lower the coordination costs, and favour the emergence of electronic markets. Electronic markets are not a side effect of the World Wide Web. Before the emergence and the wide adoption of web technologies after 1995, other forms of electronic markets did exist and were researched. The main characteristic of electronic markets is the decrease of buyer’s search costs (Bakos, 1991). This makes markets more efficient by lowering the maximum prices producers can ask. In the field of e-commerce, most price related research concerns itself with differences in prices between competitors or the flexibility of suppliers to change the prices due to lower menu costs (Brynjolfsson & Smith, 2000). With Open Access publishing, the price is fixed to zero.

Internet markets have another important effect: lower search costs lead to an increased variety of products (Brynjolfsson, Smith, & Hu, 2003). Through enhanced search and lower shelving costs for producers, more products can be made available. In the case of online bookstores, the ‘consumer welfare gain’ due to increased variety of available titles is 7 to 10 times higher than the ‘welfare gain’ due to increased competition and lower prices. The increased availability and sales of niche products is of course made famous as ‘The Long Tail’ (Anderson, 2006).

3.2.1 Open Access publishing and search costs

In the case of academic publishers – which have always functioned in a market environment – the changes in coordination costs play a very important role in the distribution of titles through Open Access publishing. Coordination costs can be further defined in several types: determining product offerings; search costs; price discovery costs and negotiation and contracting costs. Costs are not just monetary, but encompass all efforts required to achieve the result.

The costs of determining product offerings are very low with Open Access publishing. By definition, Open Access publishing grants everybody the right to access the content of a publication. The effects on search costs are the most dramatic. Making the complete content of a publication available should lower the search costs considerably, especially if search engines have complete access as well, which may lead to easier discovery of the book. When we look at price discovery costs, for Open Access books, they are very low indeed: Open Access books are freely accessible. Price discovery costs for non Open Access books are also considerably lowered, as publishers, web based book vendors and others make prices easily available through search engines. Finally, the effects on negotiation and contracting costs are not very large. For single books, those costs have never been high: there is little room for price negotiation.

Granados et al. state that ‘the more a product can be described digitally, the higher are the chances that transparent market mechanisms will prevail’ (Granados, Gupta, & Kauffman, 2006). In economics, a transparent market is defined as a market where information about offered goods is easily available to all who enter it. Making the complete content of books available enables prospective readers to examine it in detail and assess whether it fits their needs. To put it in other words: as discovering a book and reading it online is made much easier, search costs are lowered dramatically. In a transparent market, pricing information is available to all, and buyers are able to compare all available offers. Sellers have little incentive to raise their prices, but may try to find more buyers by lowering them. Lowering the prices to zero should lead to a maximum numbers of ‘buyers’.

The concept of low search costs in a transparent market perfectly describes Open Access publishing. Again, the same barriers to accessing academic publications are removed by making the complete contents available without charging money from the costumers. It was discussed before that complete access leads to better indexing by search engines, which then leads to more discovery of the publications. So, we arrive again at Hypothesis 1: The discovery of fully accessible titles is significantly higher, compared to titles which are not fully accessible. Furthermore, Open Access should also increase the usage of the publications. This leads to Hypothesis 2: The online consultation (e.g. pages read or number of downloads) of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

3.3 Resource based view of the firm

Using the theory of the resource based view of the firm and transaction costs economics, the value creating factors of e-business firms can be investigated (Amit & Zott, 2001). This is relevant for our experiment as the distribution of books published in Open Access will be done electronically, which puts an Open Access publisher in the position of an e-business firm. In order to make Open Access publishing sustainable, publishers need to create value added services for which their readers are willing to pay. Transaction costs economics have been discussed above, this paragraph describes the resource based view of the firm.

The resource based view of the firm does not consider the products, but the resources available to a firm as a means to create an optimal strategy (Wernerfelt, 1984). When the available resources are specified, the optimal market and product strategies may be found. Resources are defined as ‘anything which could be thought of as a strength or weakness of a given firm’. In more formal terms resources could be described as the tangible assets – such as machines – and the intangible assets – for instance reputation or brand name – that are available to the organization.

Using this analysis on Open Access publishing activities, several resources can be described:

- Acquisition. Attracting the best authors in the scientific fields where the publisher operates is an important part of academic publishing. Open Access publishing may be an argument for authors, as it may enhance visibility of the author’s work.

- Reputation. The reputation of a publisher could also be translated as its ranking. For a scientific author it is important to be published by a well reputed publisher. The publisher’s reputation is fuelled by already published books, series, different imprints and other factors. The use of peer review – as a form of scientific quality control – is also very important.

- Distribution. Most publishers offer worldwide distribution of publications through different partners, for both books and e-books. Furthermore, Open Access publications may also be distributed via a repository.

- Raising awareness. Publications are marketed through traditional marketing, but also via ‘new’ channels like Google Book Search.

- Technological focus is defined as an important success factor in e-business (Zhu, Kraemer, Xu, & Dedrick, 2004). A modern academic publisher is working with several technological solutions, such as the use of digital channels for disseminating publications, e-books, e-journals or other types of digital publications such as online dictionaries, blogs or wikis.

- Network. The networks in which an organisation participates may be considered an asset as well, especially as a source of new publications or innovations. Most publishers participate in several formalized and informal networks.

3.3.1 Value creation through complementaries

Several factors have been defined for the creation of value in an e-commerce environment: transaction efficiency, lock-in, novelty and complementarities (Amit & Zott, 2001). These factors will be discussed in regard to the publishing of Open Access books. Based on the results of National Academies Press, complementaries could be used to create a sustainable business model for Open Access publishers.

The first factor defined is transaction efficiency. A firm that is able to lower the transaction costs using IT creates value for its costumers. In the case of Open Access publishing, search costs can be lowered dramatically, by making the complete content of a book available both to search engines and readers. This increases the chances for readers to discover the title and – when the book is found – to assess the content directly. Here we see that both the library and information science and the theory of e-commerce point at the same mechanism.

An e-business firm can create value by motivating its customers to repeatedly use the services or goods provided. Also, value is created by creating incentives for strategic partners to maintain and improve their association to the firm. The authors name these mechanisms ‘lock-in’. In this case, the lock-in effect for its customers may be enhanced by making series of books on the same subject easily available.

Novelty is defined as a new way of conducting or aligning transactions. In the case of Open Access, the most striking novelty is making all content freely available. As discussed before, this will enhance the discovery and the usage of the publications by scientists.

The authors define complementarities as a bundle of goods which provides more value than the total value of having each of the goods separately. These complementary goods may be ‘vertical complementarities’ – extra services provided by the firms – or ‘horizontal complementarities’ that are provided by partner firms. The question is whether digital books are a complement or a replacement to paper books. Until now, sales of e-books are very low compared to the sales of paper books (Coker, 2009). This may indicate that the readers of academic books are not inclined to substitute the paper book by an electronic version. Furthermore, research by JISC shows readers dipping in and out of content rather than reading or downloading the whole book (JISC, 2009). Open Access books may function as digital representation of the paper version, by making the full contents available. While the content of a book can be accessed digitally, other ‘analogue’ properties of books are important for readers as well: paper books are more user friendly – so far. In the case of Open Access publishing, enabling the reader to directly order a paper version of the book is a complementary service. Given the low sales figures of e-books compared to paper books, this may be a valuable service for readers.

In the realm of Open Access published books, the US based National Academies Press may serve as a successful example of the positive effects on sales figures. Since 1994 all books are fully accessible online; in 2007 the number of available titles was more than 3,600. While a very small number of readers actually buy a book, due to the very large number of website visitors the organisation can sustain itself (Jensen, 2007). Based on this, it is an important question whether Open Access publishing has a positive effect on the sales of the paper versions of the publications. This leads to Hypothesis 4: The sales figures of fully accessible titles are significantly higher, compared to titles which are not fully accessible.

3.4 IT enabled innovation and dissemination channels

In the library and information sciences, the focus on Open Access publishing is explained by the impact it has on disseminating scientific knowledge. In the field of e-commerce, the main value created by an Open Access publisher is a far greater access to the published titles, which again leads to more dissemination. This should enable search engines of all kinds to index the information and thus scholars should be able to discover and assess all relevant publications for their research. The question is whether all publications are found. What measures can an academic publisher take to ensure a minimum of search costs for their readers? To be more specific: which dissemination channels are best suited to the needs of scientists?

In order to answer this question, both transaction economics theory and the resource based view on the firm provide a partial answer. Transaction economics predict whether an activity is placed within the organisation or outsourced to other organisations, depending on the production costs. The resource based view of the firm takes into account the assets of an organisation and seeks to find an optimal strategy based on that. So transaction economics emphasises the government structure of the organisation as a means to optimize transactions, while the resource based view of the firm is primarily concerned with competitive advantage using the available resources.

The differences between both theories are summarised as follows (Madhok, 2002):

Table 5.1 Transaction costs theory and Resource-based theory

According to the author, the – unique – features of the organisation must also be taken into account; as they influence how its resources interact with the transactions and its choice of governance (Madhok, 2002).

Publishing in Open Access could be described as disseminating electronic documents. Therefore, the theory on IT enabled innovation may provide a framework (Gallouj & Weinstein, 1997). It is much in line with Amit and Zott and the resourced based view of the firm by modelling services and goods through characteristics and competencies of the providing organization, its products and its clients. The authors define innovation as a change in the characteristics of one of these aspects. Adding a network of organizations working together to deliver services and goods to customers and taking into account the technology of the clients used to interface with the providers enhances this model further (Vries, 2006). This model may be useful for our experiment. A publisher cooperates with several organizations on the dissemination of its titles, by creating awareness and by distributing the titles. In other words: the publisher is part of a network that enables interested parties to become familiar with the titles offered and acquire them. In order to access the documents, the reader must use her/his own web technology to discover and read or download the content.

As described before, most academic books are in the realm of the humanities and social sciences. This implies that a large part of the readers are also active in those scientific areas. Information searching in these environments is characterised by using different information channels (Shen, 2007). A publisher may use a range of channels to disseminate its publications – most of them do not offer the complete text. Several channels offers direct access to publications: such as an institutional repository or the Google Book Search program. Both the repository and the Google Book Search program are available in an academic environment.

3.4.1 ‘Open Access’ dissemination channels and multichannel management

Firstly, this paragraph describes the characteristics of the ‘Open Access’ dissemination channels. After that – taking into account research on multichannel management – Hypothesis 5 and 6 are discussed. An institutional repository is a database used by academic institutions to store and disseminate scientific information, such as articles, books or research data. By uploading both metadata and the full electronic version of a title, a repository is used for creating awareness and dissemination. All repositories comply with the same standards, enabling other repositories that function as an information hub – such as OAISTER (OCLC, 2009) – to ‘harvest’ the metadata of the repository. Using such information hubs, scientists can find and directly access the publications and data. OAISTER is not the only information hub: in the Netherlands, all Dutch institutional repositories are also accessible though the NARCIS website (KNAW, 2009).

The Google Book Search program works in a similar way: the uploaded title and metadata are indexed and made searchable though the Google search engine. When a book is found, the user may read a percentage of the content. The percentage is set by the publisher. By setting the ‘viewing percentage’ to 100%, the content of the title is made completely accessible to the reader. The content of the publication is always indexed by Google, regardless of the viewing percentage.

Research on multichannel management suggests that customers tend to use online channels for research purposes and use offline channels – such as stores – for actually purchasing the desired goods. Furthermore, usage of multiple channels is associated with a higher purchase volume (Neslin & Shankar, 2009). This has several implications. First, allowing prospective readers full access to the content of the books may increase the number of times the books are accessed as part of the ‘customer information gathering’. This leads again to the question whether publishing books in Open Access leads to a higher discovery rate and online usage; as described in Hypothesis 1: The discovery of fully accessible titles is significantly higher, compared to titles which are not fully accessible and in Hypothesis 2: The online consultation (e.g. pages read or number of downloads) of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

Secondly, is there a difference between the discovery rate of books available through both ‘Open Access channels’ versus books available through either the AUP repository or the Google Book Search program? This leads to Hypothesis 5: The discovery of titles disseminated through both the institutional repository and the Google Book Search program is significantly higher, compared to titles disseminated through one of those channels. The same question regarding online usage leads to Hypothesis 6: The online consultation (e.g. pages read or number of downloads) of titles disseminated through both the institutional repository and the Google Book Search program is significantly higher, compared to titles disseminated through one of those channels.

Thirdly, if there is an association between availability on multiple channels and purchase volumes, does this imply that titles available through the Google Book Search program and a repository will be sold more? This leads again to Hypothesis 4: The sales figures of fully accessible titles are significantly higher, compared to titles which are not fully accessible. Furthermore, if availability through multiple channels has a profound effect, this leads to Hypothesis 7: The sales figures of titles disseminated through both the institutional repository and the Google Book Search program is significantly higher, compared to titles disseminated through one of those channels.

3.5 Conceptual model

Based on the theories discussed before, it is possible to create a conceptual model describing the impact of Open Access publishing of books on science. The role of the publisher and the reader may be depicted using the following process flow chart:

Figure 1 Process flow chart

When a scientific document is ready for publication, the publisher decides if it is published in Open Access or not. For this discussion it is not relevant whether the publisher chooses to allow the author to disseminate a digital copy or use the ‘golden road’ of Open Access publishing. Depending on this decision, the publication is only made available through the ‘normal’ channels or the publication is made available through the normal channels in combination with the Open Access channels. If the document is published directly in Open Access, the publisher disseminates the document; if the ‘green road’ is used the author will disseminate the document. The reader uses these channels to discover and assess the publication. If it suits his/her needs, the publication is read, downloaded, purchased and/or cited. Citation does not depend on the purchase of the book as the reader may access it online or borrow it via a library etc. The scientific publication process is discussed in more detail in several articles (Bjork & Hedlund, 2004; Søndergaard, Andersen, & Hjørland, 2003).

3.5.1 Constructs and hypotheses

As stated before, lower search costs enhance the usage of scientific books. The concept of search costs is defined using the construct Discovery. Discovery describes the discovery of a book by a scientist; to actually learn of the existence of the book. Knowing that a book exists leads to assessment. By online assessing the content of a book, a scientist is able to decide whether it fits his or her needs. This concept is defined using the construct Online Consultation.

Open Access publishing through different channels is depicted with the construct Accessibility by channel. Making the content fully accessible influences the discovery of the books, by opening them up to every reader and search engine. Accessibility also effects assessment, by allowing scientists to directly read the content. Discovery by scientists is dependant on the search tools used. As described before, the repository infrastructure used in universities differs from the more widely used search engines.

Lower search costs – here translated into the construct Discovery – lead to a higher research impact. In the library and information sciences, the most described form of usage is citation rate. The construct Citation reflects this. There is another measure of usage. The electronic version of a book does not replace a paper version – at least not yet. This means that a paper copy must be obtained, which would directly influence the sales figures of that title. As ordering a book can be literally done in seconds, the number of sales is a useful measure in this experiment, here defined as the construct Sales.

Research on multichannel management indicates that usage of a single channel versus multiple channels also influences the purchasing behaviour of prospective customers. Therefore, the availability of Open Access titles through one or through multiple channels may influence discovery, online consultation and – of course – sales of those titles. The construct Channel Management reflects this; it is closely connected to the construct Accessibility by Channel.

Different fields of science may have different traditions concerning citation behaviour. Furthermore, other factors may influence the scientific impact of a publication. For this reason, the construct Title Properties describes intrinsic properties of the publications. These properties are: subject, date of publication, print run and the language of a publication. The citation habits of different scientific fields may vary, and the subject of the publication is used as an indication of the scientific discipline. Also, the publication date of a document could influence citations, through the longer availability of older titles. The difference between scientific disciplines and the publication date may also influence the number of books sold. Sales expectations of the publisher can be measured using the print run. The last aspect is language, which may influence discovery.

Below are listed all hypotheses, as discussed before:

- Hypothesis 1: The discovery of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

- Hypothesis 2: The online consultation (e.g. pages read or number of downloads) of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

- Hypothesis 3: The citation rate of fully accessible titles is significantly higher, compared to titles which are not fully accessible.

- Hypothesis 4: The sales figures of fully accessible titles are significantly higher, compared to titles which are not fully accessible.

- Hypothesis 5: The discovery of titles disseminated through both the institutional repository and the Google Book Search program is significantly higher, compared to titles disseminated through one of those channels.

- Hypothesis 6: The online consultation (e.g. pages read or number of downloads) of titles disseminated through both the institutional repository and the Google Book Search program is significantly higher, compared to titles disseminated through one of those channels.

- Hypothesis 7: The sales figures of titles disseminated through both the institutional repository and the Google Book Search program is significantly higher, compared to titles disseminated through one of those channels.

The relations between the constructs are depicted below:

Figure 2 Conceptual model