Should Access to Research Infrastructures Depend on Payment?
Updated: Feb 11, 2020
Last week (12-14 Sep 2018) I had the pleasure of chairing a round table discussion (Parallel session 3, World Café) at the International Conference for Research Infrastructures 2018 which was held at Hofburg Vienna - the very impressive historical complex of the former imperial residence.
The question we addressed was whether access to RI should depend on payment. This question is closely linked to principles underlying access policies or policies on the sharing (or repartition) of operation costs for research infrastructures.
I was lucky to benefit from the contribution of distinguished participants, such as the former Director General of the European Spallation Source, Rory Fitzgerald (Director General of the European Social Survey ERIC), Boris Sharkov (Deputy Director, Joint Institute for Nuclear Research) as well as Ron Dekker and Ivana Ilijašić Veršić (Director General and COO respectively, CESSDA ERIC), to name a few. Ron Dekker was also the rapporteur of the session.
The discussion centred around one or two specific points and by no means attempted to offer a comprehensive theory on the question of access to RIs. However, here are few points emerging from the discussion, which may be of interest to research infrastructures working on their access policy or designing their policy on repartition of operation costs.
Should access to RI depend on payment?
The question whether and to what extent access to RIs should be limited to researchers from paying countries only (i.e. countries that contributed to the construction and/or operation costs) depends on the characteristics of the RI, specifically:
(1) Is the RI capital intensive (i.e. are there high construction/operation costs)?
(2) Is usage of the resource provided by the RI (e.g., beamtime, data) rivalrous or non-rivalrous?
In relation to (1) above, where the construction or operation costs of the RI are high, there
must be a built-in incentive for countries to contribute to such costs, if their research community is to benefit from the RI. Otherwise, a free-rider problem will arise whereby researchers from non-contributing countries will use the RI without payment, resulting in little incentive for governments to invest in the construction and operation of RIs (unless this is done for purely philanthropic purposes or for the benefit of science in general). If the investment is high and access is free to the user community world-wide, then governments will simply wait for other countries to make the investment and send their researchers there, rather than making the investment themselves. Obviously, there will be too little investment in RIs to serve the needs of the science community.
In relation to (2) above, there is a question whether the resource offered by the Ri is rivalrous or non-rivalrous. Beamtime (used in particle accelerators) is an example of a rivalrous resource, whereby the use of the beamtime for one experiment prevents the possibility of using the same beam for another experiment at the same time. Beamtime is therefore a rivalrous resource, i.e., a limited resource that is precious and can only be allocated to a limited number of experiments per year. Data in electronic form (available for download) or broadcasting, on the other hand, is a non-rivalrous resource. The usage of the resource by one group of researchers does not diminish the ability of other groups of researchers to use it at the same time and at no additional cost for every new user.
The above suggests certain interesting insights:
· When the costs of construction and operation of the RI are high and the resource is rivalrous, the general rule should be that access to the RI is limited to researchers from paying countries only. Otherwise, there will be too little incentives for governments to invest in the construction and operation of new Ris.
· Notwithstanding the above, there is a need to balance inequality and differences in size or financial capacity among countries. For this purpose, also capital intensive Ris offering a rivalrous resource should allocate a certain percentage of their resource to ‘world class science‘, i.e., researchers from non-contributing countries. Such percentage should be however limited.
· In the case referred to above, scientific excellence should be the only criteria for granting access to researchers from non-contributing countries.
· When the RI is not capital intensive (i.e, low construction and operation costs) and the resource provided is non-rivalrous, there seem to be no reason why access should be limited to contributing countries only. Given that costs of participation in the RI is low, incentives to join the RI as member are likely to be in the form of allocation of voting rights and ability to influence decision making process, even when access is open to all.
There are several methods to repartition operation costs among members in capital intensive Ris. Contribution to operation costs could reflect the level of use by researchers from each country. In other cases, a ‘proxy‘ will be used, such as GDP.
· When GDP is used as a proxy, there is a risk that the sharing of operation costs will not be allocated fairly. For example, it would seem to be unfair that a very large country with a large GDP but with a smaller user community will have to pay a large proportion of the operation costs in order to become a member of the RI.
· In order to address such distortions when GDP is used as a proxy, it would be important to adopt create mitigating measures, for example, allowing for different levels of memberships, such as full-member, associate member or observer, with different levels of contributions linked to each category. This will allow countries to measure the level of usage by their scientific community and move gradually to full membership as the user community grows. Experience from large RIs suggests that the user community in a country grows in numbers when that country joins as an associate or a full member in the RI.
Many other observations could be made, including a distinction between proprietary and non-proprietary use of the RI, but these are left for future roundtables or posts.