Archives and Records Management Resources
- Table of Contents
- Evidential Values
- Informational Values
by T. R. Schellenberg
Bulletins of the National Archives
Number 8 (October 1956)
There are a number of reasons why we should consciously and deliberately apply the test of evidential value in the sense in which this term has been defined and why records having such value should be preserved regardless of whether there is an immediate or even a foreseeable specific use for them.
An accountable government should certainly preserve some minimum of evidence on how it was organized and how it functioned, in all its numerous and complex parts. All archivists assume that the minimum record to be kept is the record of organization and functioning and that beyond this minimum values become more debatable. By a judicious selection of various groups and series an archivist can capture in a relatively small body of records all significant facts on an agency's existence -- its patterns of action, its policies in dealing with all classes of matters, its procedures, its gross achievement.
Records containing such facts are indispensable to the government itself and to students of government. For the government they are a storehouse of administrative wisdom and experience. They are needed to give consistency and continuity to its actions. They contain precedents for policies, procedures, and the like, and can be used as a guide to public administrators in solving problems of the present that are similar to others dealt with in the past or, equally important, in avoiding past mistakes. They contain the proof of each agency's faithful stewardship of the responsibilities delegated to it and the accounting that every important public official owes to the people whom he serves. For students of public administration who wish to analyze the experiences of an agency in dealing with organizational, procedural, and policy matters, they provide the most reliable source of what actually was done.
The test of evidential value is a practical one. It involves an objective approach that the modern archivist is especially trained to take; for his training in historical methodology has taught him to look into the origin, development, and the working of human institutions and to use records for the purpose. The test is not easy, but it is definite. It will bring to view first the records on which judgments of value can be made with some degree of assurance, the degree depending upon the thoroughness with which the records have been analyzed. It can be applied by all archivists, for no archivist is likely to question that evidence of every agency's organization and functioning should be preserved. Differences of judgment will arise only as to the completeness with which such evidence should be preserved. The test of research value, on the other hand, brings to view records on which judgments are bound to differ widely.
The information obtained by an archivist in applying the test of evidential value will also serve to evaluate the significance of records from other points of view. The archivist must know how records came into being if he is to judge their value for any purpose. Public records, or, for that matter, records of any organic body, are the product of activity, and much of their meaning is dependent on their relation to the activity. If their source in an administrative unit of a government or in a particular activity is obscured, their identity and meaning are likely also to be obscured. In this respect they are unlike private manuscripts, which often have a meaning of their own without relation to their source or reference to other manuscripts in a collection.
In applying the test of evidential value the archivist is likely to preserve records that have other values as well -- records that are useful not only for the public administrator and the students of public administration, but also for the economist, sociologist, historian, and scholars generally.
Archivists of various countries have developed appraisal standards that require the preservation of records showing how public agencies were organized and conducted their business. German archivists, in particular, have been quite precise in this regard. (Footnote 3). In 1901 H. O. Meissner, head of the Prussian Privy State Archives, formulated a number of appraisal standards that have had a pronounced effect on the German archival profession. One of these is that files (in the sense of binders of documents brought together in registries) that relate to executive direction should be preserved for each organizational unit. Among the executive matters that Meissner recognized as worthy of record were the organization, direction, housing and business arrangements, and personnel of the unit. Another standard is that general files (those consisting of records on policy, procedure, and the like that have general applicability) should be preserved in the central organizational units where they originated -- that is, where they grew out of the functioning of an organizational unit -- and not at points where they were merely transmitted or received; and that the value of general files in subordinate organizational units should be determined by taking into account the activities of such units. A third standard is that records of intermediate organizational units should be preserved if they relate to the actual management of such units and not merely to their direction from above. A fourth standard is that special files of lower or subordinate organizational units should be preserved if they relate to the management of such units. And a fifth standard is that files of judicial bodies should be preserved if they relate to the substantive activities of such bodies or if they reflect the development of permanent rights and institutions, important historical episodes, political processes, or the customs and mores of past ages.
Shortly before the Second World War the Prussian Privy State Archives appointed a special commission to formulate appraisal standards. The commission was dissolved in 1940 before it succeeded in doing this, but its activities stimulated a review of the appraisal problem by German archivists. At their meeting at Gotha, Meissner emphasized the importance of a correct archival approach in appraisal work, insisting that the old conception of appraisal as a matter of intuitive or fingertip feeling was completely discredited. His standards were endorsed by H. Meinert, who emphasized that appraisals should take into account the significance of the source of archives. This should be established by considering the position of each organizational unit in the government structure, the nature of its activities, and the relation of its activities to those of superior and subordinate organizational units. Records, Meinert held, cannot be reviewed singly as isolated pieces; they must be appraised in their administrative context.
British archivists also have emphasized the importance of preserving records on how organic bodies function. Their views on appraisal were first stated fully in a memorandum issued in 1943 by the British Records Association in connection with the wartime demand for paper salvage. In a pamphlet issued later by the Public Record Office the principles of appraisal contained in this memorandum were applied to public documents. This pamphlet, entitled "Principles governing the Elimination of Ephemeral or Unimportant Documents in Public or Private Archives," discussed the principles in relation to preserving records for business purposes and research purposes. (Footnote 4). For purposes of research the British would preserve records for three "historical or general uses": (1) to show the history of the organization concerned, (2) to answer technical questions regarding its operations, and (3) to meet possible scholarly needs for the information that is incidentally or accidentally contained in the records. The first two of these uses relate to "evidential values," the third to "informational values," in the sense in which these terms are used in this bulletin.
For the first, i.e. the history of the organization concerned, the British pamphlet favors preserving records that contain sufficient evidence to show "what was the Business or other form of organization whose activities they served--how it was conducted, by whom, and with what results." It indicates that the records containing this evidence are similar to those needed for the conduct of business. These include "Minutes and other Documents which give decisions on Policy; major series of Accounts; Correspondence leading to significant activity; Muniments of Title relating to Land and Property held by the person or organization concerned; and regularly kept Registers or Memoranda of Cases, Tests or Operations, Transactions put through or Operations carried out: roughly all the Documents reflecting policy and practice, past and present, which would enable someone else, if the present staff or practitioner were wiped out, to carry on or revive the business or work." For evidentiary purposes, the selection of records may be a bit more drastic than for business purposes, however. "Very often," according to the pamphlet, "all needs are served by preserving a few key documents and representative selections from regularly kept series and from large classes of constantly recurring documents of a routine character. Specimens should be selected for their representative character as illustrating the structure of the Business rather than for any adventitious interest . . ."
For the second use, that is, to answer technical questions regarding an organization's operations, the pamphlet would preserve evidence only for organizations that belong to "a category of Institutions or Businesses whose Archives have rarely been preserved," that are themselves of "outstanding importance" in comparison with others in the same category, or that belong to "a category of Businesses etc. the general history and development of which are of outstanding importance and can only be traced by the use of collective evidence."
Thus far we have considered the thought of European archivists on the appraisal of public records from the point of view of their value in documenting the functioning of the bodies that produced them; let us now turn to the appraisal standards relating to the evidential values of the Federal records of the United States.
At the outset it is important to emphasize that appraisals of evidential values should be made on the basis of a knowledge of the entire documentation of an agency; they should not be made on a piecemeal basis. The archivist must know the significance of particular groups of records produced at various levels of organization in relation to major programs or functions. In many Federal agencies, offices at various organizational levels build up their own files, which are usually related to and often duplicate, in part at least, those of offices below or above. In the central organizations of such agencies departmental records may be related to bureau records, bureau records to divisional, and divisional to sectional. In field organizations records of regional offices may be related to those in State offices, and records of State offices to those in subordinate offices. The use of modern duplicating devices, moreover, may lead to an extensive proliferation of records in any particular office.
In reviewing the entire documentation of an agency, the archivist's decisions on which of its records he should preserve depends on a number of factors, the more important of which are embodied in the following questions:
1. Which organizational units in the central office of an agency have primary responsibility for making decisions regarding its organization, programs, policies, and procedures? Which organizational units carry on activities that are auxiliary to making such decisions? Which field officers have discretion in making such decisions? Which record series are essential to reflect such decisions?
2. To which functions of an agency do the records relate? Are they substantive functions? Which record series are essential to show how each substantive function was performed at each organizational level in both the central and field offices?
3. What supervisory and management activities are involved in administering a given function? What are the successive transactions in its execution? Which records pertain to the executive direction, as distinct from the execution of the function? To what extent are such records physically duplicated at various organizational levels? Which records summarize the successive transactions performed under the function? Which records should be preserved in exemplary form to show the work processes at the lower organizational levels?
While an archivist dealing with modern public records will have great difficulty in reducing them to manageable proportions, he will nonetheless often find that the records he wants were not produced at all. The records on important matters with which he is concerned are often not so complete as records on unimportant matters. It is a curious anomaly that the more important a matter, the less likely is a complete documentation of it to be found. While modern technology has aided the making and keeping of records in many ways, it has also made unnecessary the production of many documents that once would have become part of the record of Government action. Much that influences the development of policies and programs never makes its way into formal records. Important matters may be handled orally in conferences or by telephone, an instrument that has been referred to as the "great robber of history." (Footnote 5).
Records on important matters are often handled much less carefully while in current use than are records on unimportant matters. This lack of care is not intentional. Policy documents cannot always be identified as such when they are first created. Policies usually arise in respect to particular transactions, and so the records pertaining to them may be interfiled with others of no lasting moment on the transactions with which they were initially associated. Records on policy and procedural matters -- on general as distinct from specific matters -- are difficult to assemble, to organize into recognizable file units, and to identify in such a way that their significance will be apparent. Records of routine operations, on the other hand, are easily managed in a routine way.
The important policy documents are also difficult to schedule for retirement. Important records on policy and procedure do not become obsolete, or noncurrent, as soon as the transactions in connection with which they may have been made are completed. The policies and procedures they establish often continue in effect. And even if those policies and procedures are superseded, the records of them serve to explain and give meaning to the change. Such records are thus difficult to retire because the period of their administrative utility is difficult to establish. Records evidencing only the execution of policies and procedures, on the other hand, usually become noncurrent when action on the particular case has been completed. The termination of routine actions is usually definite and clear. Important records, moreover, are difficult to assemble for preservation in an archival institution because many of them must first be segregated from the mass of trivia in which they may have been submerged. And at the present time this segregation commonly has to be made after the records have lost their significance for current operations and their identity has become obscured, although more effective management of current records could greatly improve this situation over the years.
Let us now see more specifically what kinds of records should be preserved as evidence of organization and function.
It is obvious that records on the origins of any governmental undertaking should be preserved. These may relate to problems or conditions that led to the establishment of a Government agency, such as a decline of agricultural prices, an increase of unemployment in the automotive industry, inequities in the regulation of interstate commerce, and the like. "Important problems," as quoted by the eminent Australian historian Dr. C. E. W. Bean from a circular sent to all departments by the Prime Minister, "are often met in their simplest form in the original stage of any undertaking. Often at that stage the object of the undertaking is most clear, and the difficulties most apparent. Records as to origin of action or organization have therefore peculiar value. Where, for example, a new Department has sprung from a branch of some other Department, and that branch itself has sprung from a Departmental Committee (or even from a public movement) which tried to grapple with the relevant problems when first they arose, the story of these initial efforts often contains the most important lesson for posterity." (Footnote 6). Records that relate to problems may be in the form of investigative reports of the executive branch of the Federal Government, minutes of hearings before congressional committees, conference minutes, and memoranda and opinions of individuals. Records that relate to the actual establishment of a Government agency may consist of statutes and Executive orders as well as drafts and supporting material relating to legislative and executive action. Records that relate to its initial activities are likely to be quite scanty. In its early stages, a governmental agency normally consists only of a few persons who are concerned with planning its organizational structure and programs. Their plans may not be committed to writing at all, and, if written, may not be preserved. For at first documents -- often of the greatest significance to the early history of an agency -- are simply shoved into desk drawers, and only after the functions of the agency have become well defined are records kept systematically in files. The administrative orders and charts that initially define the structure and programs of an agency -- the early planning documents, however sketchy and perhaps inadequate in content -- should be carefully preserved.
It is equally obvious that once an agency has been established some records should be preserved on its substantive programs. An example of how such records may be selected and reduced to manageable proportions -- to less than 1 percent of the total -- is found in the work of the Records Branch of the Office of Price Administration during the Second World War. This agency, as is apparent from its title, was concerned with the control of prices and the rationing of commodities during the war period. As a basis for establishing and fixing prices it had to gather economic data on various industries, and to obtain the observance of its regulations it had to engage in an enforcement program. Its four major programs thus related to Price Control, Rationing, Accounting, and Enforcement, each of which was handled by a major organizational unit. In preserving records on these programs, the Records Branch of the agency selected certain kinds of records on each program at all administrative levels -- national, regional, district, and local -- which in their entirety contained information on every aspect of its direction and execution.
Often summary narrative accounts exist of the direction and execution of an agency's programs. These accounts may be in the form of (1) annual or other periodic reports on accomplishments or (2) agency histories. Periodic reports, which are produced by most Government agencies, are an important but an inadequate record of accomplishment. They are inadequate because they are usually very brief, touching on just the highlights of an agency's work, and because they are usually uncritical, providing little information that is unfavorable to the agency.
Agency histories, which are often produced in relation to war emergency activities, are also inadequate as a record of an agency's work, though they constitute a very valuable supplement to its official documentation. In an article in The Library Quarterly of January 1946, Dr. W. J. Wilson, an historian in the Office of Price Administration, drew an interesting analogy between summarizations of statistical data and summarizations of records of administration and operation. He found that most of the statistical data accumulated by his agency, as well as by the War Production Board, during the Second World War could be summarized in tabulations and enumerations. He stated that "unless the masses of economic data [existing in innumerable administrative and statistical forms] are summarized statistically, they are almost useless for scientific work." Similarly he thought that "unless the masses of administrative and operating files are summarized in intelligible narratives, they are almost useless for historical work." He concluded, on the basis of this analogy, which he admitted was imperfect, that such files may be destroyed (1) "if the important historical information [in them] has been extracted and has been satisfactorily presented in narrative form . . . except perhaps for certain samples or certain illustrative documents of outstanding significance," (2) if "no historical narrative is likely ever to be based on them" because of their defective or confused condition, or (3) if they are not likely to be "used rather promptly for historical purposes." But this statement goes too far. Administrative history, just as any other kind of history, cannot be written definitively or objectively. No matter how well-conceived and well-executed an historical program may be, it cannot produce histories that will serve as a substitute for the original records. Official interpretations of records may be influenced by many factors -- the bias of the writer (which is usually an important element in the writing of official history), the competency of the writer in historical synthesis, the immediacy of the writing to the matter written about, and the like. The archivist's function is to preserve the evidence on which reinterpretations can be based, not merely to preserve current official interpretations of evidence; and to preserve this evidence impartially, without bias of any sort, and as fully as public resources will permit.
Policy documents, just as the summary reports of accomplishment, should be singled out for special attention in a record retention program. The term "policy documents," in the narrow sense of its meaning, relates to the special issuances that serve to communicate staff policies and procedures to the various line offices of an agency. No rigid distinction can be made between "policy" and "procedure." In general, policies are guiding principles that indicate the course of action to be followed in various kinds of transactions while procedures give detailed instructions on the specific methods and steps to be followed in carrying out policies. The policies and procedures may relate to matters of varying degrees of importance. Regulations, for example, are of a permanent nature; other materials of an informational character such as notices are usually of a temporary or, at most, of a semipermanent nature. The directives that embody policies and procedures may he issued in various series, according to the degree of their importance, or according to the type of function to which they relate, i.e. facilitative or substantive. They may also be issued in various forms. Directives of a permanent nature are issued in the form of manuals or handbooks; while those of a temporary or semipermanent nature from the operating standpoint, intended to be periodically superseded, are normally issued in looseleaf series. Sets of all issuances should be preserved for archival purposes. They should normally be obtained at the organizational level at which they were created. They should include issuances that have been superseded as well as those currently in effect. They may include a master set of the forms developed for each of the procedures followed. Because of careless handling of records in temporary agencies it is often quite necessary to designate specific sets of serial issuances as record sets, a procedure that is important also in regular Government agencies. These record sets may include procedural, policy, organizational, and reportorial documents. Such documents are often reproduced in innumerable copies and are liberally broadcast throughout various offices. Unless a conscious attempt is made to develop record sets, stich documents often are neither accumulated nor preserved systematically.
The term "policy document" in the broadest sense of its meaning may include many papers that relate to the courses of action followed in an agency. It may include, in addition to the series of policy and procedural issuances, all kinds of records -- correspondence, minutes of conferences, staff studies, accomplishment and special reports, legal opinions and interpretations, organizational and functional charts, memoranda defining or delegating powers and responsibilities, and the like. It may include, in a word, any paper that shows the reasons why programs came into being, as well as papers that show how the programs were administered and executed.
Policy records, in the broad sense of the term, should not be regarded as a separate class of documents. No attempt should be made to bring them together into a separate collection. During the Second World War a program was developed to create a "policy documentation file" for one of the war agencies. It was planned to select policy documents (in the broad sense of the term) and to incorporate them into a separate file organized according to the Dewey-decimal system. The criteria of selection were not sufficiently well defined, nor could they have been, for they could not be made broad enough to capture all significant documents, and if they had been made broad enough they would have been largely meaningless. As a rule when individual documents are arbitrarily torn from their context, namely from the files of the organizational units that created them, they lose much of their meaning as a record of organization and function. If records are to serve as evidence of organization and function, the arrangement given them by the organizational units that created them should be maintained: and they should not be reorganized on a subject or other basis.
The records, then, that are encompassed in the term "policy documents" should be preserved so that they reflect the day-to-day work of policy-making and policy-execution in the organizational unit that produced them. They should be selected office by office in such a way that the various groups that are preserved will show how an agency was organized and how it carried out its functions. In appraising the evidential values of public records an archivist must be particularly conscious of organization, for these values largely depend on the position of the office that produced them in the administrative hierarchy of the agency. In general, the records of offices decrease in value as one descends the administrative ladder of an agency.
Most of the significant documentation of an agency's origins and programs is found in the files of "top management." These should be preserved virtually intact for the heads of executive departments and independent agencies, though they should be purged of records on house-keeping matters. Often such files should be preserved for the senior administrative officers just underneath the agency heads, such as the chiefs of bureaus, or the chiefs of organizational units that are the equivalent of bureaus, such as services and administrations in the central organization and regional and State offices in the field organization. Records of executive direction are often embodied in bureau central files. If such records are to be preserved it may be necessary to keep large quantities of rather unimportant records along with them. If records are properly classified while they are in current use this is not necessary.
The extent to which one should go down the administrative ladder to capture the significant documentation varies from agency to agency, and is generally determined by the extent to which the activities of its organizational units are disparate in character or its administrative responsibilities are decentralized. In an executive department like Commerce, for example, the various bureaus concern themselves with such matters as weather, foreign and domestic commerce, standards, and coast and geodetic surveys. These disparate matters cannot be handled centrally except in a most general way. The important records on the programs, plans, policies, and the like are thus obviously created at the bureau level. The extent to which the records of any particular officer should be preserved depends on whether he has substance rather than a mere semblance of authority, whether he actually plans and directs and administers the work of his organizational unit or is merely the communicating agent for directions from above. The records of key officers may include their correspondence files, minutes of conferences and staff meetings, official diaries (if any were kept), memoranda, directives, and various other evidences of official action.
Attached to the offices of most heads of Government agencies are a number of organizational units that are engaged in research and investigation incidental to the formulation of plans, policies, or procedures or that are engaged in handling legal problems, budgetary matters, public relations, or internal management. Research and investigative records are of undoubted importance, for they often contain the rationale of Government programs -- the reasons why they came into being and were handled as they were. They may include staff studies and special reports which analyze workloads and performance or develop plans, policies, or procedures. Even background working papers of research and investigative offices may have value and should be examined carefully. On legal matters the archivist should normally preserve the correspondence files of the chief legal officer, opinions and interpretations, memoranda of law, delegations of authority, and other documents providing background information on the legal decisions of the agency. On budget matters the archivist should normally preserve copies of the budgets submitted to the Bureau of the Budget and the House of Representatives, and related papers such as estimates of requirements and justifications.
Public relations officers are concerned chiefly with publications, which they often merely distribute, and publicity materials, which they usually create themselves. The form of such materials is not the determining factor in considering their suitability for retention in an archival institution, for books are included among the documentary materials that fall within the definition of the term "archives." Publications produced in the performance of substantive functions should, as a rule, be preserved in libraries rather than in an archival institution. This is the case with respect to bulletins, pamphlets, circulars, and other issuances produced by agencies engaged in scientific, statistical, or research activities. There are, however, exceptions to this rule. Record sets of administrative publications created by an agency that are basic to an understanding of its functioning or organization, and publications accumulated by an agency that are basic to its own policy formulation should be retained in an archival institution. Publications embodied in records relating to their creation may also be considered eligible, particularly if the records contain successive drafts of important publications that reflect substantial changes in content.
Publicity materials produced in connection with informational and promotional activities should be preserved in an archival institution rather than in libraries. They provide the main documentation of programs that some agencies must undertake to interpret their actions to the public. Publicity materials may be in the form of press and radio releases, bulletins, pamphlets, charts, posters, and the like. They are often produced in large quantities but usually disappear almost as rapidly as they are created, for they are often considered as not falling within the definition of "records." The problem with respect to such materials is that of obtaining master files of each of the items from which all duplicate copies have been eliminated. The files should be obtained at the organizational level at which they were created. Press clippings should be retained if they are necessary to record informational activities or substantive functions of an agency on which other documentary materials are inadequate, and if they are organized in such a way as to be usable. The origin of the press clippings must also be taken into account. Nonsyndicated press clippings of specialized or small newspapers or journals should be given preference over those taken from metropolitan newspapers that are readily available at the Library of Congress.
On internal management or "housekeeping" activities, such as those relating to personnel, property, supply, and travel, relatively few records need be saved for archival purposes. In evaluating certain types of such records account must be taken of the retention of related records by the General Accounting Office, the Treasury Department, and the Civil Service Commission. The value of accounting records of particular offices for a study of Federal accounting practices, for example, is affected by the work of the General Accounting Office since 1921 in progressively standardizing Government accounting systems. Before that time the records on such practices are found in the several agencies and in the commissions that investigated contemporary practices; after that time, in the files of the General Accounting Office. The value of records of particular personnel offices, similarly, is affected by the progressive standardization of personnel procedures in recent years by the Civil Service Commission. Central records on recruitment, training, promotion, retirement, and the like, are therefore adequate; records of agencies pertaining to the administration of personnel matters should be preserved only to the extent that they reflect special or distinctive activities. The procedures that are followed in handling property and supply matters are also performed pretty much the same way in all agencies, and records pertaining to them usually do not contain much evidence essential to an understanding of the functioning of a particular agency. As a rule, then, records pertaining to internal management activities that are distinctive, that deviate from the normal pattern, or that pertain to problems peculiar to an agency should be preserved; those pertaining to normal internal management activities should not.
Records pertaining to the execution of Government programs are difficult to manage from an archival point of view. These records not only have the greatest bulk; they present also the most serious problems of evaluation. The dividing line between the executive direction and the execution of Government programs is not a very sharp one. Records that evidence genuinely significant matters relating to either direction or execution have permanent value.
While a clear-cut distinction cannot be made between records relating to the detailed execution of an agency's programs and those relating to their overall direction, a difference between the two is perceptible. In a typical Government program a number of interrelated activities occur which normally relate to more detailed matters as one descends the administrative ladder. At its bottom these activities relate to the dealings of the Government with specific persons, things, or phenomena. At its top they relate to administration and policy which are reflected in summary reports of accomplishment and more general documents pertaining to such matters. Usually the evidence on an agency's program is adequate that is provided in the form of (1) summaries (statistical or narrative) of the transactions of a specific kind, (2) a selection of records on particularly significant transactions, and (3) a selection of records on transactions that are representative of all or most of the transactions of a specific kind.
The extent of documentation required on the specific transactions of an agency depends on the adequacy of its reporting system. Under an effective system, performance will be recorded in narrative and statistical reports for administrative purposes -- to evaluate progress, to formulate or revise policies and procedures, and the like. Such reports often serve as an adequate substitute for vast quantities of detailed records on routine operations. Occasionally they may take the form of histories of activities, such as the histories of the local boards of the Office of Price Administration during the Second World War. In most agencies, even badly managed ones, the patterns of activity and the accomplishments at lower administrative levels will as a rule be adequately reflected in a limited quantity of selected documents of one kind or another. Usually such activities are conducted in accordance with orders, regulations, manuals of procedure, and other directives issuing from superior offices.
In a National Archives Staff Information Paper on "The Appraisal of Current and Recent Records," Dr. G. Philip Bauer observed that "significant variations of policy, methods, or procedure and notable occurrences usually manage to get themselves relayed upward through reports, correspondence, and complaints, or else fail to get into the records of the subordinate office."
Occasionally the summary records may have to be supplemented by records of particular actions that have special significance for an agency's history. On the enforcement of price, rationing, and rent regulations of the Office of Price Administration, for example, a limited number of case files were selected for retention (1) to illustrate the application of various sanctions, both judicial and administrative, at Federal, State, and local levels; (2) to illustrate the more interesting points of law in the enforcement of such sanctions; and (3) to document outstanding events in the agency's litigatory history. One of the criteria for the selection of the case files was thus the significance of the actions to which they pertained. The initial actions taken in new agencies or new programs may also be deserving of fairly complete documentation, even at the grassroots level of operations. Similarly, actions that represent significant deviations from the norm, if not recorded at the policy level, should be reflected in records preserved in sample as evidence of policy and procedure.
Occasionally, also, the summary records may have to be supplemented by a selection of records that illustrate the pattern or norm of action. Here the emphasis is not on the unusual or significant but on the usual or normal. Actions at the lower organizational levels may be illustrated by retaining either all records of particular offices or particular records of such offices. During the Second World War, a limited number of local price and rationing hoards of the Office of Price Administration were designated as "record boards," the records of which were preserved in their entirety to illustrate how various problems were handled at the local board level. This documentation of local hoard activities is supplemented by the histories, which have already been mentioned, and by particular classes of administrative records selected from various boards; it is probably in excess of what is needed. "Even the records of a single field office preserved to exemplify the administrative processes at the lower levels," Bauer states in his paper, "are likely to prove a disappointment when they are closely examined in relation to headquarters records." Usually it is not necessary to preserve all records of particular offices; usually a few groups or series of records taken from one or more offices contain all the evidence that is needed of the norm or pattern. A few case files on how labor adjudication cases were conducted, for example, are adequate as a record of the procedures that were followed. Usually if there is any enduring interest in the individual acts of an agency it is because of the nature of these acts rather than the governmental process involved.