As of 22 April 2009 this website is 'frozen' in time — see the current IFLA websites

This old website and all of its content will stay on as archive – http://archive.ifla.org

IFLANET home - International Federation of Library 
Associations and InstitutionsAnnual 

64th IFLA Conference Logo

   64th IFLA General Conference
   August 16 - August 21, 1998


Code Number: 088-100-E
Division Number: I.
Professional Group: Library and Research Services for Parliaments
Joint Meeting with: -
Meeting Number: 100.
Simultaneous Interpretation:   No

Research and Analytical Services for National Legislatures: A Preliminary Analysis

William H. Robinson
Senior Specialist in Public Policy
U.S. Congressional Research Service
Washington DC, USA


Currently available data on legislative research and analytical services are skimpy and contradictory, largely because of incomplete definitions about what constitutes "research" in a parliamentary setting and also because of the many actors that may be engaged in the enterprise.

This paper represents a preliminary step in ascertaining more accurately the extent and location of research services operating in legislatures. It is drawn from a new survey designed to obtain broader coverage of research units that may be based outside the parliamentary library, and offers brief vignettes describing several research services to illustrate the variety of such activities and settings.


I. Introduction and Roadmap for the Paper

Information and analysis are essential for an active and effective legislature. However, there are serious obstacles to acquiring the kind of high-quality independent analysis that will permit the legislature to play its unique role in the policy process. To permit it to function effectively, the legislature creates parliamentary structures to provide the information and analysis that it can trust and that comes in a form that a body of non-experts can use to make informed decisions on complex issues. Research capacities may come in the form of parliamentary libraries, research services, documentation services, legal departments, committee staffs, or some combination of these.

While we know a great deal about Parliamentary Libraries (thanks to their status as the oldest of legislative information institutions, and to the efforts of this Section to study them over the past three decades), we know far less about the way in which in-depth research and analysis is conducted for the legislature. Part of the reason comes from the lack of sufficiently clear definitions about what we mean by "research and analysis" and part from the fact that the functions are often widely dispersed throughout the entire structure of the legislature.

In this analysis, we survey what is known about parliamentary research through the use of existing sources of information and through a supplementary survey of such institutions conducted in the Spring of 1998. While the survey was still in process at the time this paper was written, there is enough information available to offer some informed hypotheses about what exists.

In the sections that follow, we explore the following issues: the reasons that information and analysis are so important to the legislature, including the contribution of such services in permitting the legislature to play a more active role in the policymaking processes of the nation; what is known about such research and analysis services from existing sources and why that may be insufficient and misleading; some hypotheses about what does exist, and some possible patterns of organization and operation; and brief descriptions of some of the organizations that provide research and analysis to legislatures to give more concrete texture to the discussion.

II. The Importance of Information and Analysis to the Legislature

An active legislature requires vast amounts of timely information and analysis to function effectively. This requirement stems from three related needs:

The legislature is intended to be the expressive agent of the people's will in the process of making laws to govern themselves. In doing so, the legislature reflects the common sense values and mores of the community, at the same time that it is the decentralized counterweight to the top-down, expert-driven policy process of the ministries. The collapse of command economies around the world in the past decade has demonstrated the fallibility of human knowledge and the infeasibility of attempting to run everything from the center. Much of the same reasoning applies in the political realm as to economic matters. Only the legislature has the moral authority and political muscle to challenge the "one size fits all" prescriptions of a characteristically technocratic executive style, and to tailor policy to fit the needs and desires of its citizens.

III. Obstacles to Effective Legislative Research

Yet in crafting such an analytical capacity, the legislature confronts obstacles - political, economic, organizational, and cultural in nature. Regardless of whether the form of government is presidential or parliamentary in design, the executive is not enthusiastic about giving the legislature more analytical firepower that could potentially raise havoc with executive proposals. Presidential systems that vest more fiscal autonomy in the separate branches of government have the advantage in this regard, since under most parliamentary forms, the legislature is still dependent on the Treasury or Finance Minister to grant sufficient resources to create a credible analytical group.

In any event, it takes resources and competitive salaries to create a research capability, and not all nations can muster sufficient funds to do it. Deciding where to place the organization(s) is another challenge. Should it be a single unit or multiple entity? Concentration of power is one of many dimensions of this issue. Another dimension is how to ensure balance and objectivity for a group that reports directly to the powerful head of the parliamentary staff. (Oversight committees made up of Members is one device that has worked in some countries).

Finally, the legislature needs a particular kind of research - applied research that is sometimes called policy analysis. It is not always easy to find sources of analytical expertise that combine the skills needed: rapid response capability, objectivity, focus on practicality, ability to communicate in non-technical terms, in-depth research skills, issue expertise, political sensitivity, and knowledge of how the legislature operates. The frustration of attempting to engage part-time academics in time-bound practical analysis is well known in the social science literature. This has driven many legislatures to create their own research capacities so that the legislature can capture the institutional memory which this engenders.

IV. Current State of Knowledge About Parliamentary Research Services

Most of what we know about parliamentary research services comes from a survey of parliamentary libraries done in 1995-96 (reported in the World Directory of National Parliamentary Libraries, edited by Dr. Ernst Kohl and published by the German Bundestag in 1996). While that survey was directed at parliamentary libraries, there were three questions addressed to research for the legislature for the first time: "number of research and subject analyst staff in the Library", whether or not "analysis and research" was being conducted ("by the Library", or "by which other department or organization"), and "any other information (including information on other Parliamentary documentation and information services which do not come under the Parliamentary Library)." Despite a valiant attempt, the effort was not wholly successful in drawing forth the real dimensions of research and analysis undertaken by parliaments. The reasons are (1) the questionnaire and subsequent reference volume was so focused on parliamentary libraries, that it was impossible to lift attention to a broader set of institutions; (2) there was no definition of what was intended by the terms "research and analysis," and (3) since the questionnaire was sent to libraries to complete, it appears that those respondents had become so accustomed to a historically narrower set of questions that they really did not break free and respond in expected ways to the new questions on research. It is clear that many thought the questions on research only applied to the library (despite the careful wording of the questions themselves).

Given the pattern of responses, we can be most confident about the information from the Directory when research is conducted by the library or one of its component units-- which is true in only about half of the usable cases (and only one-quarter of the total cases in the Directory). Moreover, the responses on numbers of staff can only be used for parliamentary libraries, since the question clearly asked for the number of "researchers and subject specialists" in the Library. Finally, there were a number of ambiguous responses and a large number of non-responses in key areas. For example, even for those libraries stating that they conducted research, only slightly more than half filled in the number of research staff. Moreover, the ambiguity about what constitutes "research and analysis" plagues the questionnaire even in its strongest possible uses. Several chambers answered "no" to the question on research and then proceeded to list the number of researchers in their library.

Other questionable responses were more serious. A few examples will suffice:

Having noted the limitations, what can be said about our knowledge of legislative research at this point? First, it is necessary to determine what usable data remain after some necessary adjustments. The Directory lists 203 legislative bodies-- with a few multiple entries for countries with two chambers. Of these, 9 nations had no parliament, 21 more had no parliamentary library, and 7 did not respond to the questionnaire-- leaving 166 usable responses. However, another 53 chambers are not likely to have research services (37 that answered "no" on research and were not likely from their scale or political culture to find a research function essential; along with another 16 chambers that did not answer the question on research, but were unlikely from their scale and/or history to support a research operation). There were 17 chambers that possibly had research services but could not be finally determined from the evidence offered in the Directory (8 that answered "no" on research but their scale and political culture made it likely that they might have a research capacity; and another 9 that did not answer the question, but were possible based on scale and/or political culture-- or were actually known to have unreported research capacities). This leaves the 98 that answered "yes" to research, and about which we can speak with some level of confidence-- provided we do not push the data too far.

First, as to location of research, the results are almost equally divided between research being based in the library (50 respondents), and research being conducted by another organization in the legislature (48 chambers). There is no significant difference in this relationship by region of the world - with Europe, Latin America, and Asia and the Pacific about equally divided between these two patterns. (Only Africa showed a distinct 2-1 preference for locating research in libraries as opposed to separate research organizations). For all usable replies where research is done outside of the library, about half (24) had a separate research unit. The balance spread research responsibilities among legal departments (5), documentation services (4), and a residual category made up of standing committee staff, archives, other parliamentary staff, the Secretariat, or multiple sources (12).

Second, in terms of scale, legislative research is likely to take place in relatively small organizational settings, if it follows the pattern of parliamentary libraries. Parliamentary libraries had median staffing levels of 4-5 in 1995 in large areas of the world, including Central and Eastern Europe, Asia and the Pacific, and the Near East and Africa. The median scale is 10 for Latin America and the Caribbean, and 15 for Western Europe. Since they are generally younger in origin, and followed the longer-existing parliamentary libraries, most research groups are likely to be smaller-- if they are not already a smaller component of the library.

Looking at the Directory questions on scale, we find some confirmation of that suspicion. There were 34 chambers that answered "yes" to research, sourced that research in the library, and entered the number of researchers. Of these, two-thirds had five or fewer staff allocated to research. For those libraries answering "yes" to research, but not responding to the question on the number of staff, 76% had total library staff of ten or less-- making it likely that no more than 5 would be researchers. Looking at slightly larger organizations (those in the range of 6-10 staff for both groups), picks up 8 more units-- meaning that 37 of the 50 library-based research units had 10 or fewer staff. There are a very few large units that skew the average. There are 4 organizations with 31-40 research staff (Chinese Taiwan, Egypt, Italy, and the U.K. House of Commons), and another group numbering 50-100 research staff (Australia, Canada, and India), with the U.S. Congressional Research Service (with 444 research specialists) as the largest. A few more large ones may be uncovered in a follow-up survey now being conducted by the author.

Based on earlier research that I have conducted on this question, I suspect that the larger parliamentary research groups will be arrayed roughly as follows (leaving room for inevitable surprises): CRS with 444 researchers (all numbers are for 1995-96); the Legislative Advisory Department of the Brazilian Chamber of Deputies next with 188; Russia (with over 150 research staff divided between its two chambers); the Japanese Research and Legislative Reference Bureau (152); the Reference and Research Services Directorate General of the German Bundestag (123.5 research staff); Poland (with 108 researchers divided between the two chambers); India (93); the Research Service of the Italian Chamber of Deputies (70); Canada (80); Australia (43); the People's Assembly of Egypt (36); Argentina (36); and the British House of Commons (34). The numbers may all change somewhat, but the relative ranking is likely to be very similar to the listing above.

V. Some Early Returns About Research Services Around the World

While more definitive information must await the returns from the new supplementary questionnaire that was sent to Secretaries General (to deal with some of the problems uncovered in the World Directory), there are some interesting patterns already evident.

Parliamentary research services are highly variable. While parliamentary libraries have settled into some clearly discernible patterns regarding kinds of services offered, recognizable organizational frameworks, and methods of operation, the newer research activities are more eclectic, and come in a rich array of organizational settings and offer a wide variety of different services. To provide some cohesion to the discussion, I suggest a typology that is based on the degree of integration of services. At one end of the continuum are the integrated services - where research and reference activities are located in the same agency (often the library), and the two services are well coordinated. Next come articulated services, where library and research are part of a broader organizational framework, but the two activities of research and reference are only loosely joined. Separate services form the next point on the continuum - where two separate organizations must reach accord to meld research and reference activities. The dispersed services model occurs close to the other end of the continuum. In this model, even research services are dispersed among several organizations or are offered from separate disciplinary perspectives. The final point on the continuum is the absence of research services-- which puts every Member and every committee on its own devices to gather necessary information. There are advantages and disadvantages to each model; at this point in the analysis, there is no attempt to judge (explicitly) one pattern as superior to another.

This analysis was done in a personal capacity by the author and does not necessarily represent the views of the U.S. Congressional Research Service.