Talk in open access circles of late has centred around the true costs of setting up and maintaining an institutional repository. The only accurate answer to this question, in my opinion, is: it depends - on a number of factors.
At the low end of the cost range is the completely free institutional repository. An individual can easily download free software, such as gnu eprints, using computing and internet facilities already in place for other purposes at work. The amount of volunteer labor involved also depends on how the IR is set up. Are authors allowed to deposit their own works, or is there a central vetting process? Obviously, the central vetting process is more labor-intensive than allowing authors to deposit their own works.
Of course, even this is option is not truly completely free; it is just that there are no hard dollar costs. Even with no budget at all, we can easily get an institutional repository up and running with what we have.
In fact, this might be easier and simpler for the smaller and poorer library. Decisions, for example, are easier, when one has fewer options to contemplate. This is another example of the Delightful Irony of open access; that the poor can afford, what the rich cannot (or claim that they cannot).
At the higher end of the cost range, a large university could plan a comprehensive institutional repository program, not only for the open access research literature, but also for all manner of other types of information. For example, universities which move to electronic records only have long-term needs for preservation and access to a variety of institutional records, including some (such as student and personnel records) which are confidential. It may well be necessary and/or desirable to develop a series of repositories, rather than just one, and these may be handled in a centralized or distributed manner. The highest single per-repository cost would come with a central system housing a variety of different types of information for a large university. This operation may well require a fair bit of hardware, connectivity, security and authentication arrangements, staff, and space to house the computers and staff - plus adminstrative overhead, of course. The costs for this kind of arrangement are not necessarily new costs, as there will be current means of maintaining institutional records and so forth; indeed, there may be a variety of cost savings. It is likely that there would be initial transitional costs, such as equipment purchases, space design, and so forth. Even with substantial cost savings, however, the costs for such a comprehensive program could well be substantial, both initially and ongoing.
In between there is a great variety of possibilities for costs. A library could set up an institutional repository using free open source software, and advertise for a specialist to work on the institutional repository, whether as technical support or promotion or both, full or part time. An open source software solution with paid hosting costs could be pursued, or commercial software.
To sum up, when we look at the wide variety of costs reported for institutional repositories - from practically nothing to $6,000, to hundreds of thousands of dollars, and ask: which of these costs estimates is correct? There are two correct answers to this question: all of the above, and it depends - how much money do you have, and what are you willing to spend?
This wide range of costs is not particularly unusual, especially in the electronic environment. Think of a website - or blog. Either one can easily be set up by an individual, using equipment and connectivity already in place as part of a standard internet connection service - or even using totally free equipment and services, often provided by public libraries. Or, a large institution can set up a web services department, hire a manager, some graphic designers, administrative staff, and naturally, supply them with office space. Here, too, it is quite accurate to say that the true cost of a website or blog ranges from nothing or virtually nothing, to a great deal of money.
This phenomenon is not entirely new, for that matter. Consider the costs of transportation. A billionaire might own a fleet of planes, helicopters, ships, and fancy cars, in various locations all the over the world. Transportion costs for such a person could be astronomical. Those of us with more modest means get about with our bikes, skateboards, more modest cars, and buses. Even the destitute, barring disability, can get about on their own two feet. If we think such a person cannot get as far - remember Gandhi. Has anyone accomplished more? So, the answer to the question: how much does transportation for one person cost? Is, quite correctly, a wide range, from nothing to a very large sum of money, with many possibilities in between.
This post was inspired by a recent conversation on the SPARC Open Access Forum (SOAF); for more on this perspective, see my SOAF posting on this topic.
This post reflects my personal opinion only and does not represent the opinions or policy of the BC Electronic Library Network or the Simon Fraser University Library.