Conclusion: Digitization and the Continuities of Change in Administrative Information Processing

Introduction

When we look at administration, the notion of information is ubiquitous. In a sense, administration is all about information processing: gathering, storing, searching, reusing, reinterpreting, and sharing information in a never-ending communication chain or, rather, in an ever-expanding communication network.

From the early chancellery administration based on information organized chronologically in minutes, through the modern administration with its files organized by systems of registration and reference numbers supplemented by card indexes, to the present-day administration using elaborate digital information management systems integrating all this in seamlessly interconnected databases, there has been constant change not only in the way information is processed but also in the form and content of the information itself (Gugerli, 2012; Vismann, 2000). This undoubtedly has had and will have far-reaching consequences that go beyond the field of administrative practice. At the highest political level, too, one is aware of this link between the way we work in the office and the way society is governed. After her election as President of the Swiss Confederation for 2015, Federal Councilor and Head of the Federal Department of Justice Simonetta Sommaruga puts it in a nutshell. In an interview with a renowned Swiss newspaper she said that “in the past we used to send a letter; today we send an e-mail and then a text message. This leaves marks in politics” (Sommaruga, 2014).

This statement refers to the two fundamental aspects of administration: its internal modus operandi and its external interdependencies with government, society, and the economy. Accordingly, Jos C. N. Raadschelders (2009) distinguishes two approaches that administrative history can take. “Administrative history proper deals with administration itself: tasks, organization and functioning, and actors”; “administrative history in the broader sense,” on the contrary, would focus on “the wider background,” “thus the relationship between citizen and government, the processes of state-making and nation building, and developments in international relations” (p. xiii). A similar distinction is echoed in organization theory between a micro and a macro perspective. While microlevel analyses focus on the way single organizations adapt to their individual task environments, macrolevel approaches deal with aggregates of organizations and the characteristics of organizational sectors (McKinley & Mone, 2003).

In this conclusion to the symposium, I begin with some conceptual and theoretical considerations that help to clarify the central terms: administration, process, information, media, and power. The intention is not to provide precise definitions, but to suggest a framework that is flexible, yet coherent enough to integrate the diverse perspectives presented by the contributors. Bearing this framework in mind, I take up the motifs of continuity/discontinuity and trust/mistrust that, in some form or other, have appeared recurrently throughout this. Along these lines, I conclude with a concise review of the papers, situating them within the broad field of interaction spanned by the materiality of hybrid administrative processes and the ideological dynamics in society.

Administration

Students of administrative history can draw on the writings of many notable authors who cover a wide range of different aspects that are worthy of scrutiny (Adler, 2009; Pugh & Hickson, 2007; Raadschelders, Wagenaar, Rutgers, & Overeem, 2000; Sager & Rosser, 2009; Shafritz, Ott, & Jang, 2005; Tijsterman & Overeem, 2008). Max Weber (1921-1922/1980, pp. 551-579) saw the modern bureaucracy as the instrument of rational domination. Woodrow Wilson (1887) discussed the politics–administration dichotomy. Herbert Simon (1947/1997) analyzed administrative behavior and decision-making processes in hierarchical organizations. Niklas Luhmann (1971) approached them as social systems of communication.

These few, deliberately chosen, examples suffice to illustrate how manifold the study of administrations and its history is. Implicitly building on this great tradition of theoretical writings on administration, the contributions collected in this volume begin from a more pragmatic starting point: the manifest process of rapidly progressing digitization of administration and its potential to change not only the administrative organizations and workflows themselves but also the way states and business enterprises administer the rights, duties, wishes, and preferences of citizens and customers.

Because of this pragmatic approach, I refrain here from attempting to provide an abstract, generally valid definition of what administration—and in particular public administration—is (Raadschelders, 2009, p. 7). As for the 19th and 20th centuries I can, however, indicate some central characteristics of these conglomerates studied as administration: They are organizations with a compartmentalized, usually hierarchically structured architecture based on specialization and a clearly defined division of competences. They work according to well defined, rule-based procedures or programs that imply the close interlocking of humans and technical devices. Their main raw material is information in any form and their main product is information in the form of performative decisions. The basic medium of this transformation of information is the dossier as part of a registration system.

However, as dramatic as the effects of digitization may appear today in light of big data and artificial intelligence, and the growing capacity of states and businesses to supervise and cater to ever more transparent citizens/consumers, digitization has not arrived overnight. New information and communication technologies (ICTs) began making important inroads into administrative practice during the second half of the 19th century. Since the 1970s, historiography began to study these secular changes from a diachronic perspective. In his pioneering study “The Visible Hand: The Managerial Revolution in American Business,” Alfred D. Chandler (1977) traced the emergence of the modern multiunit business enterprise that fundamentally transformed the U.S. economy in the decades between 1840 and 1920. He argued that these hierarchically organized enterprises resulted from a trend that gave administrative coordination an advantage over market mechanisms in securing greater productivity, lower costs, and higher profits (for a transaction cost perspective on this issue, see, for example, da Silva & Saes, 2007; Ouchi, 1980; Williamson, 1991 and, for a critical view, Simon, 1995). Although Chandler focused on organizational rather than technical development, his study indicated that the economic advantage of administrative coordination depended on technological innovations, above all the railroad, the telegraph, and the telephone. He certainly opened the door for further studies focused on the relations between information technology and (administrative) organization (Beniger, 1986; Cortada, 2011; Douglas, 1986; Graber, 2003; Yates, 1985, 1989, 2005). Indeed, the growing interest in studying information from a historical perspective led the information historian Toni Weller (2011, p. 2) to speak of an “information turn” in scholarly thinking.

The anthology “A Nation Transformed by Information. How Information Has Shaped the United States from Colonial Times to the Present” edited by Alfred D. Chandler and James W. Cortada (2000) marked a milestone in this history. It made a case for the importance of information in its many forms and technical manifestations as a critical factor in the development of social, economic, and political structures. Above all, the efforts to rationalize industrial production that gained momentum around 1900 attracted the attention of many scholars. For example, in her study “Structuring the Information Age” on the introduction of punched cards and later computers in the U.S. insurance industry, JoAnne Yates (2005) stressed the path dependency of the introduction of new technologies in private administrations and showed how users in practice influenced new technologies; in her words, “ . . . users influence not just the technological artifact that vendors sell but also ‘technology-in-practice,’ or what users do with the technology” (Yates, 2005, p. 2).

Even before the economically driven rationalization by ICT in the private sector, public administrations adopted new ICT to cope with ever-expanding information needs, a secular process dramatically accelerated in the 20th century by World Wars I and II. In his seminal study of the British civil service in the 19th and 20th centuries, for example, Jon Agar (2003) shows how ICT affected and changed administrative practice and organization. Briefly summarized, he argues that the growing information needs of the state and influential expert movements championing new technologies were the main drivers in the increasing mechanization of bureaucratic procedures. But this process was full of setbacks and by no means linear; there were, at many points along the way, powerful political and administrative interest groups that managed to frustrate change.

While this is the familiar way in which we look at the issue today, Agar also suggested an inverse perspective. He argued that metaphorical ideas about administrative practice influenced the design of new ICTs. In this view, the conceptualization of administration as a machine was a prerequisite for the mechanization of administrative processes. The “discursive mechanization” created the ideals of how proper administrative practice should work, and these in turn shaped the way mechanical devices were designed.

One illustrative example of this was the “Analytical Engine” devised by the English mathematician, philosopher, inventor, mechanical engineer, and would-be politician Charles Babbage in the 1820s/1830s. With this concept, he materialized the idea of a strict separation between thought and labor, which was the guiding principle of organization in the civil service at the time. According to this metaphorical construction of the administration as a machine, the “Analytical Engine” was to be controlled by instructions stored in the form of holes punched into cards, making direct human intervention in the operational working process obsolete (Agar, 2003, pp. 39-69). With reference to digital ICT, this metaphorical approach is echoed in the Memex idea by Vannevar Bush, one of the later masterminds of what became the Internet. In his article “As we may think,” Bush (1945) sketched out a device that, imitating the network structure of human thinking, would work as a “Memory Extender.” Designing the device, Bush fell back on the notion of the office. Memex had the form of a writing desk with microfilm as a storage medium. It enabled projection, registration, and cross-linking of documents. It was never built, but with its fundamental concept Memex came to be considered as the precursor of the computer and the origin of technically networked documentation and communication (Rapp & Bender, 2013).

The metaphorical proximity between administration and ICT is no coincidence. It rests upon an essential inner connection. The core business of modern administration, as it developed during the second half of the 19th century, consists in the controlled processing of information according to predefined bureaucratic “programs.” Information is both the most important asset and the most important product of administrative action. It has been subject both to the (power-mediated) formal and informal rules of the administrative organization, and to the resources and limitations of the (culturally mediated) technologies for recording, cataloging, storing, and (re)using information in practice. In this sense, recent developments in ICT can be embedded in a long-term perspective that examines the mutual interdependencies of ICT on one hand, and administrative principles and practices on the other.

Hybrid Processes

Given the current dominance of digital information technology in administration, such a long-term perspective highlights an issue that might otherwise remain unnoticed. The spread of the computer has introduced a new distinction within the administration, one that separates a “digital” from an “analog” world. Mechanization had never led to such a separation of mechanical and non-mechanical “worlds.” Mechanization considered administration to be a unity of human and machine. Administrative processes were still thought of as conducted and controlled by humans, partly supported by machines. The reference point remained the known administrative practice, and the question was how mechanical devices could make this practice more efficient.

While, at the beginning, digitization schemes were conceived in similar ways, they gradually began to emancipate themselves from the analog administrative processes and create their own reality (Gugerli, 2018). Now the question was not how technology would support administrative processes, but how these processes could be replaced by digital information processing, and ultimately how they could be automated. In media theory, the new performative potential of “the electrical” in contrast to mechanical procedures had already been recognized by Marshall McLuhan (1964/1996, pp. 347-359) in the 1960s. Yet considerable parts of bureaucratic processes have remained analog to this day, provoking constant border crossings between analog and digital. This creates a hybrid field of action with multilayered transitions that require a constant reevaluation of the interface between human and machine.

There were three main obstacles on the way to full automation. First, input to digital processes usually comes from human beings; second, the output of these processes is directed toward human beings; and third, depending on the available data and technology, human control is needed at certain (decision) points in the process (Machlup, 1983, p. 660). So, administrative processes assumed a genuinely hybrid character with more or less numerous passages from digital to analog procedures.

I will briefly illustrate this hybridity with a common administrative process of our own archival practice: the process of ordering archival files. Not so long ago, this was entirely analog. The client had to fill in a form and hand it in at the desk. Then, an archivist would check whether the file was legally accessible and confirm this by signing the form. The form consisted of two sheets: a control slip and a carbon copy. Once accessibility had been confirmed, another member of staff would fetch the box containing the file from the repository and hand it over to the client, placing the carbon copy of the order form as a surrogate where the box had been in the repository. The control slip would be stored in a card box in the back office of the reading room. At one time, the boxes given out to each client were manually registered in a database to keep a record of the executed orders. However, this practice turned out to be too costly and was soon abandoned.

A first step in digitizing this process was made possible by the implementation of the digital online catalog. Now, the client ordered the files directly via the shopping cart of the digital catalog. The order was registered within the central archival information system. But then the process became analog again. The data in the system were printed on an order sheet consisting of two parts: a delivery slip for the client and a surrogate for the repository. In obvious cases, according to the available metadata and predefined rules in the database, the system marked the form as “free,” so only the remaining forms had still to be signed by an archivist to confirm accessibility (see Figure 1).

OPEN IN VIEWER

Figure 1.

Swiss Federal Archives: Schematic hybrid ordering process.

This simple example shows a typically hybrid administrative process going back and forth between analog and digital information processing. Pushing the digital frontier further by substituting digital information processing for analog subprocesses brought two administrative benefits above all. All orders were now registered automatically and the archives had, at no extra cost, complete control of clients’ orders at any time. At the same time, the semiautomation of the subprocess of authorizing access improved the efficiency not only of this subprocess but of the process as a whole. The “free” order sheets could be passed directly from the printer to the logistics staff, thus eliminating two process interfaces by omitting the loop passing via an archivist.

In a next step of digitization, the subprocess of authorization was integrated into the database. Now the archivist authorizes the orders by ticking checkboxes on the screen. This makes the whole process more efficient, as the digital–analog break in the middle is eliminated and it is only at the end that the process becomes analog by printing the order form. It is easy to think of further automation of the ordering process. For digital files that are freely accessible, a fully automated process from order to delivery without any human intervention is technically feasible already today. In the near future, the machine will be able to grant personalized access to files still subject to a legal closure period.

It might be interesting to look into this kind of phenomenon more closely. How did ICT systems and administrative practice deal with this hybridity? What were the conditions for and effects of further automation pushing back analog information processing? How was this changing hybridity connected with administration’s outside interaction with the public and other institutions? How did it and will it, eventually, affect the relationship between state and citizens? How far can automation be driven anyhow? Is it conceivable to eliminate hybridity altogether and bring Agar’s metaphor of the Government Machine to full fruition? or Will there remain residues of analog processes? and Is there even something like an optimum point of hybridity?

Information

The contributions to this special issue could not clarify all of these issues. We will have to leave this to further studies. By taking the perspective of information history the present papers do, however, shed light on various broader subjects connected to the interdependencies between ICT and administration. But what is “information”? In his paper, James Cortada suggests understanding information as a tool, like a physical object. Marian Adolf and Nico Stehr, on the contrary, stress the peculiar character of information. Like ideas and knowledge, its transmission from a producer/sender to a customer/receiver does not imply a transfer from one to the other, as the sender remains in possession of the sent information. Setting off “information” against “knowledge,” Adolf and Stehr consider information as merely reflecting attributes of objects, whereas “knowledge” “has an enabling quality giving actors the capacity to act” (see also Stehr, 2003, pp. 42-49). These two notions seem to reflect the fundamental conceptual distinction between “information as an object” and “information as a sign” (Capurro & Hjørland, 2003). In fact, the ubiquity of information defies any single clear-cut definition. With the library and information scientist Michael K. Buckland, “[we] conclude that we are unable to say confidently of anything that it could not be information” (Buckland, 1991, p. 50; see also Luhmann, 1998, pp. 1091-1092). For this reason, we need a pragmatic understanding of “information,” one that sets the concept in relation to a specific context (Capurro, 2009; Capurro & Hjørland, 2003). In the administrative context, I would like to take up the argument of Erhard Oeser (1976, p. 86) that information is a “system-relative concept.” Toni Weller corroborates this view from a historical perspective. Questioning the usefulness of rigorous definitions of information, she asserts that “such definitions are not particularly useful to the historical study of information, since information should be defined and understood in relation to the historical context, which changes” (Weller, 2008, p. 17).

As discussed above, administrative processes, in some way or other, always involve humans and machines. Conceiving this hybrid configuration as a sociotechnical system, we could distinguish a social from a technical subsystem. The social subsystem is constituted by meaning accessible to humans. The sociologist Niklas Luhmann (1987, pp. 193-194) pointed to the key role played by information in the ongoing production of meaning. Meaning, he argued, emerges from the processing of differences that require a selection between options. It is information that connects the differences and enables selection and, ultimately, understanding. In this view, information appears rather as an object in the form of a meaning offer suggested by the sender and decided upon by the receiver. Consequently, the information is not the same for sender and receiver. In fact, information is only established within communication processes (Capurro & Hjørland, 2003).

The technical subsystem of administration does not necessarily depend on meaning and lends itself to more technically oriented concepts that consider information mainly as signals transmitted from sender to receiver that can be somehow measured. The “Mathematical Model of Communication” by Claude E. Shannon is a paradigmatic example of these transition models. In its engineering context, meaning becomes completely obsolete (Capurro & Hjørland, 2003; D. Chandler, 1994; Shannon, 1948). However, as stated above, this technical subsystem too must, in the end, produce meaningful results in hybrid processes. An interesting approach to bridging the gap between information as (meaningless) signal and information as meaning offer is the concept advanced by Benny Karpatschof (2000, pp. 131-132), who conceived information as a low-energy signal that triggers a release mechanism that, in turn, can cause high-energy reactions. This apparently mechanistic concept is open for meaning because it draws attention to the subjective condition of the release mechanism that accounts for the selection of distinctive options (Capurro & Hjørland, 2003).

Of course, these remarks do not provide an overarching definition of information, not even for this symposium. But they emphasize the context sensitivity of the concept of information, delimit roughly the semantic field it covers, and offer some clues about how to deal with its broadness.

Media

Taking information as an object in the context of communication or information flow in the manner of Luhmann or Karpatschof brings into focus its material aspect. This is particularly true in administrative contexts. Administrative information is typically put in written form and conveyed through and stored on preferably reliable and stable carrier media, and it follows predetermined procedural rules. In the analog world, ideally, the main administrative carrier medium for information is paper documents organized in files properly registered within a filing plan (Vismann, 2000). The information, or the signs that make up the information, is inscribed on the carrier medium. Mechanization brought recurrent changes to this setting of information, media, and inscription. Compared with the formerly handwritten documents, the typewriter accelerated writing, eased reading and copying, and required new kinds of ink and paper. Punched card machines depended on a special type of solid paper and format. They introduced a form of inscription that was readable very fast for machines, but no more directly decipherable by people (Agar, 2003, pp. 149-155). Xerox copying and further mechanical copying techniques eased the exact replication of inscriptions from an original document onto many paper copies, boosting distribution and circulation of information (Dommann, 2014, pp. 241-252; Yates, 1987).

Digitization brought first the central mainframe computer, mainly used for statistical calculation, then the dispersed personal computer supporting specific office tasks. Later, these were integrated into local networks and, eventually, into the Internet (Agar, 2003, pp. 333-387, 417-432; Cortada, 2004/2008, 2009, 2012). This network integration constituted the dominant information and communication technique of administrative practice, and its main carrier medium for information became the relational database (Gugerli, 2009, 2012). All this accentuated some of the changes in the setting of information, media, and inscription already under way while making others obsolete. But digitization also added new qualitative dimensions to these changes. Information in digital form could be processed much faster and in amounts unthinkable before. Compared with analog retrieval instruments, such as card indexes, databases offered new possibilities for information retrieval and filtering, facilitating, even automating the detection of patterns—a cumbersome task of counting and calculating in the old days. Information in digital form was not only easier to retrieve, communicate, distribute, copy, access, and analyze but it was also more malleable (Mussell, 2013). In fact, this is one of the central advantages of digital over analog information. It not only allows the easy correction of errors but also facilitates its adaption, combination, and aggregation according to evolving information needs.

But all this comes at a price. The same advantages of digital information pose new challenges regarding security, authenticity, integrity, data protection, and reliability. Sebastian Gießmann demonstrated in this symposium what is needed to make digital information trustworthy. He argued that it was (and still is) only by the complex interlinking of machine-based and (human) administrative procedures that trustworthiness was (and is) generated. In particular, he makes two claims: In the perspective of media theory, trust, and with it intersubjective reality, is constructed by these “chains of mediation.” From an information perspective, trust is based on authenticity of the relevant information, or, at least, the actors’ belief in it. Generalizing from Gießmann’s thesis, I argue that the construction of authenticity depends on many factors, but it changes with different media and the ways information is inscribed on them. To understand this, Bruno Latour’s concept of immutable mobiles is helpful (Latour, 1990, 1992). Latour looks at how—in what form—information is passed between actors, or, generally, between actants, which could be human or nonhuman (Latour, 2005). Immutable mobiles are objects of particular characteristics. They have to be mobile, that is, they can easily be moved in time and space. At the same time they have to be immutable, that is, they are stable and do not change no matter what movement they are exposed to. Latour (1992, p. 21) itemizes further characteristics of these immutable mobiles: presentability, readability, and combinability. They make immutable mobiles valuable tools that give their users decisive advantages in the pursuit of convincing and mobilizing others and, thus, build up power networks.

Latour developed the concept in the context of his studies on modern scientific culture. In particular, he was looking for an explanation of its origin and of the specific traits that made it so powerful. While the laboratory was the paradigmatic location of Latour’s study of science, he explicitly mentioned the use of immutable mobiles in bureaucratic, that is, administrative contexts. The statement is so significant that it is worth citing at length:

The connective quality of written traces is still more visible in the most despised of all ethnographic objects: the file or the record. The “rationalization” granted to bureaucracy since Hegel and Weber has been attributed by mistake to the “mind” of (Prussian) bureaucrats. It is all in the files themselves. A bureau is, in many ways, and more and more every year, a small laboratory in which many elements can be connected together just because their scale and nature has been averaged out: legal texts, specifications, standards, payrolls, maps, surveys . . . The “cracy” of bureaucracy is mysterious and hard to study, but the “bureau” is something that can be empirically studied, and that explains, because of its structure, why some power is given to an average mind just by looking at files: domains that are far apart become literally inches apart; domains that are convoluted and hidden, become flat; thousands of occurrences can be looked at synoptically. More importantly, once files start being gathered everywhere to insure some two-way circulation of immutable mobiles, they can be arrayed in cascade: files of files can be generated and this process can be continued until a few men consider millions as if they were in the palms of their hands. Common sense ironically makes fun of these “gratte papiers” and “paper shufflers,” and often wonders what all this “red tape” is for; but the same question should be asked of the rest of science and technology. In our cultures “paper shuffling” is the source of an essential power, that constantly escapes attention since its materiality is ignored. (Latour, 1992, p. 26)

Power

In our context, Latour’s concept does three things. (a) It helps us to follow the paths of information in concrete settings on a microlevel and to analyze the many tiny interferences and passages between different practices of inscription. (b) Looking conceptually at these micro processes is key to understanding the media-dependent changes of inscriptions and their “social” effects over time. (c) Eventually, the concept of immutable mobiles allows the micro to be connected with the macrolevel (Blok & Jensen, 2011, p. 124; Latour, 1991, p. 118; Latour, 2005, p. 213ff). This last point is important, because it provides a clue to new perspectives that enable us to connect analytically the inner mechanics of administration and its outward interactions in the wider political and social context. In particular, it guides us in analyzing questions of power and domination in their cultural-material ramifications in the daily practice of actors knitting networks that connect humans, nonhumans, machines, and things. It is within such knitting processes that power differentials accrue and domination is stabilized. Following the actors in their efforts to unite (human and nonhuman) allies to sustain their cause, Latour detects “macro-structures” (Latour, 1991, p. 118). In the context of science studies, he denotes them “centers of calculation” (Latour, 1992, p. 29). For general social contexts, Latour (with reference to Jeremy Bentham’s “panopticon”) coined the neologism “oligopticon” for this phenomenon of network configurations that constitute powerful macro-actors maintaining extensive actor-networks (Blok & Jensen, 2011, p. 121f). But their power rests in concrete, localized contexts that are shaped by the “ongoing circulation of information between the oligopticon and its many points of connection” (Blok & Jensen, 2011, p. 122).

It is important to underscore that the power relations in these actor-networks are of a sociotechnical nature and constitute material arrangements assembling humans and “things” like machines, bodies, and texts (Law, 1991, p. 186). Modern, bureaucratic administrations, in particular public administrations, have often been characterized metaphorically as machines (Agar, 2003; Collin & Lutterbeck, 2009). In Latourian terms, they can be considered as hybrid macro-actants with manifold connections with a hybrid world. In many aspects, Latour’s procedural focus at the same time enlarges and adds detail to the observations of important writers on bureaucracy such as Max Weber and Niklas Luhmann. For Weber (1921-1922/1980, pp. 551-552), the rule-bound, factual execution of businesses was a central element in the technical superiority of modern bureaucracy. Weber’s rigid ideal-typical notion of a prefigured process was opened up to contingency by Luhmann’s system theoretical conception of social procedures that order decision processes in a specific functional perspective, but whose concrete course depends on the selective decisions of the participants who progressively eliminate alternative courses and, thereby, reduce complexity. The outcome of the procedure is the result of this contingent cascade of decisions (Luhmann, 1969/1997, pp. 201-218). The medium of such administrative procedures is the file, which in the course of the procedure assembles the documents (texts, graphs, tables, photos) containing the arguments for and against different possible solutions (Nellen, 2011, p. 11).

It is interesting to note that for Luhmann (1997) the legitimacy of legislative, administrative, and judicial procedures resides precisely in their procedural mode, because their decisions cannot be anchored in self-evident scientific truth. This clear-cut contrast between a scientific and social notion of truth is subtly qualified by Latour’s own anthropological study of laboratory life, where he demonstrates the processes of aggregating and filtering of information by means of immutable mobiles that produce scientific facts. While for Latour, the laboratory remains a point of reference as a “truth-production site” in Western civilization, he has endeavored to study other such sites such as the Conseil d’État (Council of State), the supreme court of administrative law in France, where he demonstrates “the Laboratory Life, not for the construction of facts, but for the construction of legal arguments” (Blok & Jensen, 2011, p. 117; Latour, 2010, p. ix, and particularly Chapter 5 on scientific objects and legal objectivity; Nellen, 2016).

So, drawing Luhmann and Latour together, we could aim to scrutinize the “Laboratory Life” of public administration as a site for the construction of “binding decisions” (Luhmann, 1971, p. 182). We could also include administrations of the private sector, although we should rather think of them as sites of binding contracts. Both forms, public and private, constitute key institutions of modernity. Both follow a mode of operation of bureaucratic hierarchy that contrasts sharply with modern ideals such as democracy, the market, and freedom. Max Weber (1921-1922/1980, p. 835; Weber, 1920/1986, p. 203) aptly dubbed this uneasy mode of operation “stahlhartes Gehäuse”—translated quite freely into English by Talcott Parsons as “iron cage” (Clegg & Lounsbury, 2009, pp. 118-119; Kaesler, 2012). However, it would be wrong to consider the “iron cage” as merely a complement to emancipatory principles, a necessary evil, so to speak (Kieser, 2006, pp. 76-78). While in industrialized countries the tune of the “excessive bureaucracy” is standard, the dire effects of rampant corruption and failed states in many regions of the globe indicate that the “iron cage” of the bureaucratic mode is a prerequisite for modern societies to function (see also Cook, 2018).

State Simplifications

It is in this broader context of modernity or, if anything, multiple modernities (Eisenstadt, 2000) that we should heed Latour’s (2010) admonition “that we can’t possibly provide a positive anthropology of the Moderns . . . as long as we don’t have a clear comparative study of the various ways in which the central institutions of our cultures produce truth” (p. ix). This special issue has aimed to contribute to this agenda by way of a multiperspective study of the history, present, and future of administration as the central modern institution for the production of power-mediated “truths.” The approach allows for a great variety of contributions: Case studies digging deep into the micro mechanics of administrative practice, and interpretative analyses evaluating macrostructural effects and potentials of new technologies. The diverse perspectives are held together by a concept largely guided by the ideas of Bruno Latour’s actor-network theory and his reflections on the role of inscriptions and immutable mobiles within institutional truth-production. It provides an analytical framework for an analysis of the interdependencies between the materiality of internal administrative processes and their external interdependencies with other “centers of calculation” in society, the economy, and politics by conceptualizing administration as a hybrid field of intense social and technical interaction rather than a social system or as a social actor within an environment or a social context.

This anthology started out from the idea that public administrations are the “information machines” of states. As such, they are a key element in the study of the intricate relationship between state and society. In his seminal history on the “Sources of Social Power,” Michael Mann (1995, p. 61) noted two secular trends in modern societies. On one hand, modern, bureaucratically organized states had “infinitely more control over” their citizens than premodern states. At the same time, conversely, “civil society’s capacity to control the state also increased.” Due to this mutual interpenetration, it becomes unclear where the state ends and where civil society begins:

The state is no longer a small, private central place and elite with its own rationality. “It” contains multiple institutions and tentacles sprawling from the center through its territories, even sometimes through transnational space. Conversely, civil society also becomes far more politicized than in the past, sending out diverse raiding parties—pressure groups and political parties—into the various places of the state, as well as outflanking it transnationally . . . The “power” of the modern state principally concerns . . . a tightening state-society relation, caging social relations over the national rather than the local-regional or transnational terrain, thus politicizing and geopoliticizing far more of social life than had earlier states. (Mann, 1995, p. 61)

The interpenetration of state and society can take many forms, but most of the time administration is involved, and it does not seem exaggerated to claim that the daily business of state–society interpenetration is realized through the information infrastructures of public administrations; or, as Max Weber (1921-1922/1980) put it, “In daily routine, domination is first and foremost: administration [Herrschaft ist im Alltag primär: Verwaltung]” (p. 126). Although Weber focused solely on state power and did not yet consider the penetration of the state by social interests, he was well aware of the central role of administration within the state–society relation.

To enable the state to act upon and interact with society, its information machines need to have a minimum knowledge of relevant features of its counterpart. But as social reality is always much more complex than can be grasped by any administrative information processing, states have to simplify. James C. Scott (1998) studied the fatal consequences of this mechanism in his seminal study “Seeing Like a State” on the failure of despotic, high-modernist development projects in Third World nations and Eastern Europe. Considering state simplifications “the basic givens of modern statecraft,” he argues that only by simplifying and standardizing can a state “make a society legible” (Scott, 1998, p. 2). Scott compares these state simplifications to abridged maps:

They did not successfully represent the actual activity of the society they depicted, nor were they intended to; they represented only that slice of it that interested the official observer. They were, moreover, not just maps. Rather, they were maps that, when allied with state power, would enable much of the reality they depicted to be remade. (Scott, 1998, p. 3)

It is important to note here that Scott did not solely blame the state’s attempts at “rationalizing and standardizing what was a social hieroglyph into a legible and administratively more convenient format” (Scott, 1998, p. 3) for the disastrous outcomes of the utopian social engineering schemes analyzed in his book. Administrative ordering of nature and society is a necessary, but not sufficient condition for disaster. According to Scott, three more elements have to be present: a high-modernist ideology, an authoritarian state, and a prostrate civil society (Scott, 1998, pp. 4-5). By themselves, state simplifications are ambiguous in their consequences:

[T]hey are as vital to the maintenance of our welfare and freedom as they are to the designs of a would-be modern despot. They undergird the concept of citizenship and the provision of social welfare just as they might undergird a policy of rounding up undesirable minorities. (Scott, 1998, p. 4)

Administration and Society

These considerations may suffice as evidence for the close link between the details of administrative information processing and broader historical processes. Against this backdrop, the case studies by Frohman on Germany and Brugger and Koller on Switzerland shed light on the intricacies of state registration practices at the point of transition from analog to digital information processing. This is particularly interesting because it points to the continuities and discontinuities of administrative information processing and qualifies the many ideological claims of a “digital revolution.” The ideal of a universal integrated register of the population is not an invention of the digital age, and the causes of the recurrent failures to put the ideal into practice are not just technical. Thus, it is essential to know the history of information processing to assess the potential of digitization and where its effects may disrupt or stabilize existing social structures. Taking up the assertion by Swiss Federal Councilor Sommaruga cited at the beginning, it is true that the shift from paper to digital formats leaves marks in politics and affects the state–society relationship, but the way in which this is taking place and the directions changes are taking are by no means linear and straightforward (see, for example, Gugerli, 2018). It is therefore a good strategy to look first for continuities and gradual changes before identifying revolutionary ruptures. However, paradigmatic shifts do occur. When, as noted by Frohman, the implementation of integrated information systems “transforms information processing from a bureaucratic burden to a political resource,” we might have found one of these revolutionary moments.

An important notion in the digitization of administrative information processing is “trust.” Governments and administrations trust in the potential of new technologies to generate more trustworthy information on society and nature more rapidly, thereby leading to better decisions. At the same time, these technologies trigger mistrust and anxiety against totalitarian control schemes and the loss of individual privacy. Koller noted that data collection by the state depends on citizens trusting that it will handle the information they provided correctly. While differences exist to this effect between different regime types, it is in principle true even for authoritarian and despotic states, as citizens might withhold or distort information wherever possible (Scott, 1985).

In order fully to make use of the potential of ICT, state registries need a unique identifier for each individual that is to be registered. As discussed both by Frohman and Brugger, this raises severe questions of trust. The example of the long-standing discussion on the use of the Swiss Old Age and Survivors’ Insurance (OASI, AHV in German) number as a unique identifier is particularly interesting as it shows how, for a long time, it was administrative obstinacy rather than privacy concerns that frustrated such plans. Privacy concerns became prevalent only at a late stage, when criticism of state handling of personal data and anxieties about the safety of ICT emerged. Recently, a risk analysis on this very subject commissioned by the Federal Office of Justice and the Federal Data Protection and Information Commissioner stated that, in 2017, more than 14,000 administrative and organizational registers used the OASI number as a unique identifier (Basin, 2017). It warned,

The associated computer systems . . . are subject to attacks, from both internal and external attackers, resulting in unauthorized access to the register data. This risk is not negligible as many of the systems that store and process these registers are maintained by organizations, such as municipality administrations, schools, and hospitals, that are not subject to the same high security requirements as federal IT systems. (Basin, 2017, p. 5)

Data protection and privacy are usually addressed in terms of regulations that would instill trust in digital information processing by harnessing the risks of abuse and totalitarian control while preserving its advantages of customer convenience, and business effectiveness and efficiency. But trust is also generated within administrative procedures themselves. In his paper on the development of digital payment practices, Gießmann shows how trust is generated by implementing hybrid administrative procedures involving humans and bureaucratic media technologies in sequences of identification, de-identification, and reidentification. His approach of interpreting these procedures as part of a continuous world project in registration and identification that is known today as the “Internet of Things,” but that harks back to the era of industrialization of the 19th century, is a good starting point for further long-term studies of digitization. They could, for example, tie in directly with his reflections on the strong drive toward standardization, certainly one of the most important features implied by digitization. For these processes of abstracting “information from its original context” (Frohman) may point to larger dynamics of a changing state–citizen relationship. Together, the case studies of Gießmann, Frohman, Koller, and Brugger point to the need for comparative and transnational approaches to bureaucratic registration practices as a crucial element within more and more globally entwined “projects of legibility” (Scott, 1998) by states and private corporations.

Gießmann’s attention to underlying continuities is mirrored in the critical assessment of the allegedly revolutionary potential of big data technologies by Adolf and Stehr. Establishing a continuity between the 19th-century concept of Social Physics and the 21st-century visions of the feasibility of social engineering by thoroughgoing informationalization of all aspects of life allows them to qualify both anxiety and enthusiasm regarding the dangers and promises of big data. Focusing on the dangers, Adolf and Stehr point to a series of issues that scholars of administration and information history should consider, such as asymmetries regarding the capability of accessing and processing information or questions concerning the control of person-based data. They also point out that digital information environments have little tolerance for ambiguity—a reformulation of the strong standardization bias of digitization.

However, it is not the immense technical potential of big data that first and foremost alarms Adolf and Stehr. More importantly, they criticize the far-reaching ambitions of the proponents of digitally enhanced “Social Physics” on the grounds of ideology. The claim of being in possession of the means to overcome, eventually, the technical limits that have frustrated so far all projects of total information collection is a promise that may turn “Social Physics” into a powerful ideology for a future society of total transparency. Citing Robert Merton’s assertion that such a society of total visibility would be a “fiendish” one, Adolf and Stehr consider ideologies of this kind a real danger for a liberal democratic society. It will remain to future studies sensitive to the interplay of continuities and discontinuities to examine whether the current promises of big data and artificial intelligence may change “the basic givens of modern statecraft,” and free states and administrations from the need for simplification. And it will be for the future to reveal whether this is for good or ill.

The notion of continuity is also emphasized by James Cortada, whose paper opened the discussions in this volume. It is therefore no coincidence that he starts his narrative of the interdependencies between ICTs and administration in 1967, when digital technologies first began to be embraced by firms and state bureaucracies. In this perspective, the term revolution has to be used with utmost caution when describing the entwined histories of ICT and administration. Rather, they are characterized by a dynamic of incremental, evolutionary changes (recall the Swiss Federal Archives’ ordering process). Cortada suggests three elements that are key to the study of these changes: technology, processes, and information. Each of them influences human and mechanical activities that are interconnected by information. While technological developments are relatively well known, the gradual transformation of processes in organizations, and the way information itself was changing, have been less investigated. Cortada places particular emphasis on information, as it is this element that can serve as a conceptual framework for interdisciplinary approaches that focus on integrating all three elements. With the notion of information ecosystems, he also offers a methodological concept for such analyses. Information ecosystems are characterized by mutual dependencies among the actors, a strong responsiveness of actors to information, and a certain predictability of information–(re)action chains that allow the actors to build up reliable expectations.

Information ecosystems can be useful for heuristic purposes and for the definition and delimitation of meaningful objects of investigation. By mapping the patterns of information within these systems, the relations between its elements and their changes over time can then be traced. Depending on the concrete issues under scrutiny, this endeavor can and should draw on a variety of concepts and theoretical approaches, and adopt different perspectives. But the considerations by Bruno Latour on actor-networks, inscriptions, and immutable mobiles seem particularly well suited to guiding studies of the sociotechnical environments, and their complex hybrid configurations of analog and digital processes under consideration here.

Starting from the discussions at the Bern conference and summarizing the contributions to this volume, this conclusion suggested some concepts that could be helpful for administrative history focusing on information. Each paper tried to take a step in this direction. All of them provide answers to some questions and open up a number of interesting new ones that remain to be answered. It is with this in mind that this symposium developed: to suggest some lines of research for scholars interested in the history of administration and information. In the light of the ongoing debates about information and disinformation, the use and misuse of information, totalitarian surveillance, and the role of ICT in all this, their studies will be more relevant than ever, to help keep the information society informed.