WHO OWNS CONTRACT AND GRANT DATA IN THE U.K. AND WHO CAN USE IT? BY Cally Brown School of Library Archive and Information Studies University College London II Marcia Taylor I SSRC Data Archive I" University of Essex In 1980, a working group of the Social Research Association published a report on the "Terms and conditions of social research funding in Britain". (1) Issues singled out for special consideration included control over publication and owner- ship of data and copyright. The ensuant discussion was not as useful as it might have been, however, oecause the differences inherent in contract-funded research and grant-funded research were not always recognised. It therefore seems worth- while here to start by distinguishing between these two concepts insofar as they relate to the British situation. f In grant-funded research, money is awarded by the commissioning body on a broad understanding as to the results of the research. The study is usually initiated by the researcher rather than by the funding body, and the funder does not normally see itself as the primary user or beneficiary of the results. Rather, the intended audience is seen to be other practitioners and theoreticians in the general field of the enquiry and, ultimately, 'the citizen'. Ownership of results-- including both data collected and interpretations of those data--are usually left in the hands of the researcher. In contract-funded research, money is awarded to the researcher for a specific study defined by the funder. The researcher may be pre-selected by a 'closed-tender' process, or chosen from a group invited to apply for the contract on a competitive 'open-tender' basis or, occasionally, appointed after public advertisement. A customer-contractor relationship is entered into where the commissioner purchases the researcher's services and is the prime user of the research results. The funder usually retains much clearer control over the research process than in the case of grant-awarding bodies, and usually claims rights of ownership over all material created during the activity of the enquiry. In effect, the contracted researcher is paid for his time and expertise but has no rights to the product of his labour. As the Social Research Association report suggests, public funding of social research in the U.K. is organised in two ways: through government departments and quangos, and through more generally oriented independent bodies such as the Social Science Research Council (SSRC). Increasingly, government departments are using contracts to administer research, whilst the SSRC usually allocates research funds through a grant system. The report also suggests that central government departments are increasing their direct support of social research whilst, at the same time, public funds made available for more general social research dre being diminished. With this picture in mind, we have first looked at the ways in which central government departments directly administer their research funds, paying special consideration to contract practice and how this affects ownership of, and access to, data; and then we have gone on to consider Britain's most prolific independent funding agency of social research, the SSRC. The examples used as illustration have generally been based on personal communication. PRACTICE IN CENTRAL GOVERNMENT DEPARTMENTS r ,1 Funding Conditions The extent of control exercised by government departments over externally conducted research initially depends on whether the funds have been awarded on a contract basis or on a grant basis. Practice varies considerable Those departments that usually employ a grant system for funding research do not, typically, have a tradition of internal research activity. This may be due to a variety of factors, but in particular, may be due to the nature of the policy area involved. For example, the Department of Education and Science and the Department of Health and Social Security are concerned with the school system and The National Health Service respectively. In both instances, the policy control is decentralised, and in both cases, the department favours a grant system as the most appropriate way of administering research funds. Departments with a tradition of performing their own research and which have-- or have had--internal research units, are most likely to work on a contract basis. The Department of Employment, the Department of the Environment and the Home Office, for example, all tend to put researchers under contract. Access to Research Results As we have already suggested, contract-funding can give the funder greater control over the research he has sponsored. Under common law in the U.K., an employer has the right to claim ownership of all materials created by an employee during the period of employment, so long as this is stipulated in the employment contract. The employee's right to publish no longer applies although under the Copyright Act ownership of copyright arises from evidence of authorship. For example, in British universities, ownership of copyright material authored by researchers is usually claimed by the university in its employment contract. Whether the university subsequently exercises this power of ownership is another question. Similarly, government departments can claim ownership of work performed by externally contracted researchers. 10 iii Her Majesty's Stationary Office, apparently, advises departments to include a specific clause in contracts making any written results of research subject to Crown Copyright, and the inclusion of a further clause claiming ownership of materials encoded in machine-readable forms. (Ownership of machine-readable data is particularly unclear in English law as this area is not addressed by the 1956 Copyright Act currently in force.) The inclusion of such ownership clauses in government contracts is, however, left to the discretion of individual civil servants. This means that not only does contract practice vary between departments, but it can also differ within a department. In some contracts, the most restrictive veto on publication is both stated and implemented. In others, the department reserves the right to prior publica- tion--indeed, in one case where such a right was reserved, the sponsoring depart- ment used this right to delay publication of results, which meant that the researcher concerned was unable to publish his somewhat contradictory interpreta- tion . The publication veto is, however, usually more liberally interpreted by indivi- dual civil servants and although the formal contract appears to be restrictive, the researcher will often receive an accompanying letter of agreement relaxing any publication restrictions stipulated in the contract. It can also be noted that contracts from a few departments place no conditions upon publication apart from requiring acknowledgment of sponsorship and a waiver of departmental responsibility. The variation in contract implementation practice within departments is perhaps best illustrated by two views independently expressed to us concerning the same department. One researcher stated that the department was "the best--very liberal" in its attitudes, whilst the other was emphatic that the department was "a bugger-- always gives me trouble". Machine-readable data are preserved, if at all, on the initiative of individual government departments or by the Social Survey Division of the Office of Population Censuses and Surveys. There is, at present, no Public Records Office machine- readable archive where records collected in this form by government research must be deposited. Although in March 1981 a review committee recommended to the Lord Chancellor that such an archive be established with the greatest possible speed (2), the future development of such an archive remains a matter for conjecture and discussion continue with the SSRC Data Archive as to the possible form this might take . Meanwhile, public access to government-initiated data remains discretionary. Typically, data are made available through individual arrangement between researchers and civil servants. More general access is frequently provided through deposit in the SSRC Data Archive and localised access is sometimes made possible by deposit in one of the smaller data archives established in several British universities. In the absence of legislation, access is often as haphazard as the interpreta- tion of publication rights. Some examples may serve to illustrate this. 11 In one type of departmental contract, the data are provided by the department to the contracted researcher. In one such case, the researcher was obliged to sign The Official Secrets Act and was forbidden to allow anyone access to the data who was not specifically named in the contract. He was further required to return the data uncopied. In another case where the department supplied the data, however, the researcher was permitted to mount a copy on his local machine in perpetuity. Another contract reported to us stipulated that the data be destroyed or returned to the departments after use but constructed variables could be retained by the researcher. An absolute lack of caution on the part of the department is illustrated by one somewhat incredible case where the researcher was supplied with a sample of highly confidential, individual records taken from a central register. Nowhere in the contract was reference made to the preservation of confidentiality or to subse- quent use of the data. In the other type of departmental research contract where data are collected by the contracted researcher, there are equally contrasting examples of depart- mental attitudes towards data access. In one case, a seemingly non-controversial enquiry in the area of medical research, the researcher, himself, was very keen to deposit resulting data in the SSRC Data Archive. The funding body, however, remained adamant that the data should be withheld, effectively ensuring that no further access could be made to the data by the contracted researcher or secondary analysts. In direct contrast to this, a well established research institute which con- ducts numerous social surveys of medical care under government contract, deposits data as a matter of course in the Data Archive. The lack of any legislation defining public rights of access to data may have been a contributory factor in the last example where data were scheduled for deposit in the Data Archive following a project initiated during the life of one government but, with a change in administration, this decision was reversed and no further access was granted. As we see it, where social research is administered directly by government departments on a contract basis, the government funds the research, specifies the research and has the power to control dissemination of the research results. There is, however, no consistency in practice and ownership and access conditions vary between contracts. Current Developments There is no automatic right of access by U.K. citizens to public records. Access is not governed by any written rule of law but is, as we said earlier, at the discretion of the government and, de facto , of individual civil servants. There is, however, an indication that this situation may be changing and that the government is becoming aware of the need for some sort of coherent policy on public access to government data and--by extension--to government contracted re- search data. 12 ' In 1980, a review of the Government Statistical Services was carried out under the chairmanship of Sir Derek Rayner. It recommended, among other things, that government departments should seek less costly and more flexible means of enabling interested members of the public to have access to government figures, and that clear rules about the use of data should be published in order to enable more statistical research to be performed outside the civil service. (3) Where this recommendation is acted upon, it may well help to increase the public availability of data collected under government contract. An indication of this already happening can be seen by the number of government department approaches made to the SSRC Data Archive in order to use the Archive's facilities for disseminating data and statistical series, PRACTICE IN THE SOCIAL SCIENCE RESEARCH COUNCIL Funding Conditions The Social Science Research Council (SSRC) was established in 1965 to promote and fund research activities within the social sciences and to provide a continuing overview. The bulk of its research funds are spent on grants. The grant-awarding process is entirely reponsi ve--that is, application is made to the Council at the initiative of the individual researcher. Proposals are reviewed by a committee appointed for this purpose by the Council, consisting of leading scholars in the field, and then put out for independent refereeing. This 'peer group assessment' is considered to be an essential part of the award process. Awards are typically made to institutions rather than to individuals, so that applications are subject to further scrutiny by the research committess in appli- cants' home institutions. Access to Research Results Once a grant has been awarded, the researcher is usually left to his own devices to complete the research and submit his report. A copy of this report is usually deposited by the SSRC in The British Library Lending Division. Additional publica- tions by the researcher are encouraged by the Council, and copyright rests with the investigator. The SSRC aims to assure access to any machine-readable data generated during the study by making it a condition of the grant that a copy of the data be offered to the SSRC Data Archive for subsequent use by secondary analysts. Failure to do so may affect the success of any future grant application made by the researcher. Current Developments In recent years, responding to both a growing shortage of funds and to public pressure to make research more relevant to policy issues, the SSRC has allocated an increasing proportion of its grant budget to specific research initiatives de- fined by a specially appointed board of the Council. 13 Administration of these research initiatives more closely approximates the contracting methods used by government departments, where the procedures are more formalised and supervision is likely to be more stringent. Whilst directing research may be an efficient way for the SSRC to administer its restricted resources, this trend towards a research-initiatives policy has led to an emphasis being placed on short-term, ad-hoc and specific policy-oriented research at the expense of more long-term, basic research. In spite of the SSRC's moves towards initiating and directing research, there has been increasing criticism of the Council for supporting esoteric and irrelevant studies, culminating in demands in parliament and the press for its closure. (4) However, the Rothschild Report published in May 1982, outlinging the results of a review of the SSRC's functions and functioning, recommended that the Council should not only not be closed, but that it should be asked to return more dili- gently to its original remit of promoting the future development of social science research, particularly mul tidiscipl inary research that will not only advance the understanding of current issues of public importance, but will also fundamentally question the working of society. (5) The report stressed the importance of 'peer review' of social research, em- phasising the need for independence from government departments in research initia- tion. If the recommendations in this report are implemented, the main thrust of SSRC funding can be expected to return to a grant-awarding system with its more liberal copyright and data access arrangements. It can reasonably be expected that independence from government control of data access will be assured by the Council's continued committment to the broadest possible airing of research results. Concl usion In the volatile and often contradictory situation we have described, it is difficult to envisage specific suggestions which could be made to guarantee the public availability of government-funded data. The Social Research Association has begun drawing up a list of "desirable and undesirable contract conditions", and it recommends that "steps be taken to secure agreement to such a list". (1) Whilst we feel that the first of these tasks is formidible and the second - monumental, nevertheless we feel that such an exercise is a necessary pre-requisite to any further action. We also echo the Association's view that many of the difficulties surrounding ownership of, and access to, data would be "alleviated if there were greater har- monisation of contract condi tions" . (1 ) We would add, in conclusion, that until data protection regulations and free- dom of information legislation are in force in the U.K., little progress can be made towards any consistent or just policies to ensure access to data. 14 NOTES AND REFERENCES (1) Social Research Association, Terms and conditions of social research in Britain . (London: SRA, 1980) [report of a working group] (2) Modern public records: selection and access . (London: HMSO 1981) [report of a committee appointed by the Lord Chancellor. Chairman - Sir Duncan Wilson, March 1981.] Cmnd. 8204 (3) Rayner, Sir Derek, Review of government statistical services . (London: HMSO 1981") [report to the Prime Minister, December 1980] (4) See, for example. The Guardian 20 May, 1982. (5) An enquiry into the Social Science Research Council by Lord Rothschild . (London: HMSO 1982) [presented to parliament in May 1982] Cmnd. 8554 1982 ANNUAL CONFERENCE REPORT The lASSIST 1982 Annual Conference and Workshops were held May 27-30, at the Hotel del Coronado in San Diego, California, U.S.A. According to only moderately biased reports, the conference was considered a huge success by those attending. The site was fantastic, the program participants were well prepared and the hospi- tality suite closet bar was a different experience. Conference attendance highlights include: - Ninety-five participants over the four day period. - Good workshop participation. Fifty-three percent of those attending the conference attended a workshop. - Significant Canadian, European and Australian representation. Over 22% of the participants were not from the United States. - Sporadic participation by the hotel police force in the post mid-night session of the Turkey Action Group. The police comprised about 10% of the group. - Amazing representation at Sunday's business meeting. Thirty-seven per- cent or 35 people appeared. Approximately 50% of the conference participants were not previously lASSIST mem- bers. Twenty-four or 25% of those individuals subsequently became members. Only 44% of the conference participants were from California. 15