Thinking about efficiency in libraries

Citation

Bade, D. (2010), "Thinking about efficiency in libraries", Journal of Documentation, Vol. 66 No. 1. https://doi.org/10.1108/jd.2010.27866aae.001

Thinking about efficiency in libraries

Article Type: Comparative review From: Journal of Documentation, Volume 66, Issue 1

Efficiency and Management,Guy Callender,Routledge,London,2009,xiii, 303 pages,ISBN: 9780415431804

The ETTO Principle: Efficiency-Thoroughness Trade-Off, or Why Things That Go Right Sometimes Go Wrong,Erik Hollnagel,Ashgate,Aldershot,2009,162 pages,ISBN: 9780754676775 (hbk)

Effizienz statt Gerechtigkeit? Auf der Suche nach den philosophischen Grundlagen der Ökonomischen Analyse des Rechts,Klaus Mathis,Duncker & Humblot,Berlin,2009,255 pages,ISBN: 978-3-428-12724-5

Efficiency instead of Justice? Searching for the Philosophical Foundations of the Economic Analysis of Law,English translation by Deborah Shannon,Springer Verlag,New York, NY,2009,220 pages,ISBN: 9781402097973

The Myth of Resource Efficiency: The Jevons Paradox,John M. Polimeni, Kozo Mayumi, Mario Giampietro and Blake Alcott,Earthscan,London,2009,200 pages,ISBN: 9781844078134

Keywords: Library and information networks, Resource efficiency, Library management

Reading the final report of the Library of Congress Working Group on the Future of Bibliographic Control I could not help but notice that efficiency as a value that trumped all others was evident throughout the report. In Bade (2009) I remarked “The demand for efficiency begs the question of what we are trying to do in our activities of bibliographic control, i.e. efficient at doing what?” (Bade, 2009, p. 1). This question prompted me to set out to look more broadly and more carefully at the discourse of efficiency in LIS. One of the first items I located was a 1913 essay by a cataloger at the John Crerar Library named Aksel G.S. Josephson. He offered the following definition of efficiency:

• •

  Efficiency exists when we achieve the desired results with the least expenditure of effort and the smallest amount of resultant waste … When our task is accomplished in such a manner that every step counts directly toward it, when every effort results in a corresponding accomplishment, when the amount of resultant waste is a minimum, then we have before us the results of efficiency (Josephson, 1912-1913, pp. 7-8).

Josephson put this in the context of a library by asking whether or not “the present bibliographical condition” is efficient. In more concrete terms he asked “Can we gather together anywhere in this country a complete series of bibliographical records of any subject that anyone wishes to investigate? Are the existing and available records adequate? Is the material referred to in the records available?” (Josephson, p. 9). While his apparent desire for “a complete series of bibliographical records of any subject” ought to raise plenty of objections, for the purpose of investigating our understanding of efficiency the passage – including this assumption – is instructive. Anyone familiar with the library literature of the late twentieth and early twenty-first century will note the sharp contrast between Josephson’s association of efficiency with completeness, adequacy and availability from the viewpoint of one using the library, and the reduction of efficiency to matters of cost for the library that characterizes the literature of our time.

In the library management literature efficiency has come to mean economic efficiency, often simply cost-cutting, and has frequently been discussed as something separate from effectiveness. There have been voices of caution such as Hernon and McClure (1990, p. 8) who wrote at the beginning of their book:

• •

  Attention to efficiency rather than effectiveness fails to consider the likely tradeoff relationships between the two types of measures. Efficiency of an activity or service cannot be improved past a certain “critical point” without injuring the effectiveness of that activity or service, and vice-versa. Thus, continued attention to increasing the efficiency of various activities and services may become dysfunctional or counterproductive to effectiveness criteria.

More recently Marc Storms made the same distinction in his complaint about “too many managers running around libraries in The Netherlands”:

• •

  Management that focuses only on making sure that everything keeps running efficiently is not a path I would recommend. Everything needs to work effectively … Efficiency means the right cost and employing the right means, but effectiveness means getting the job done (Storms, 2003, p. 9)

Unfortunately the majority of references to efficiency in the literature of library management are not so carefully considered, and nowhere have I found any detailed discussion of the nature of efficiency. Efficiency is either defined to fit the writer’s purposes or simply assumed to be unproblematic. The literature on efficiency in relation to information technologies suffers from the same kind of treatment.

Outside the boundaries of LIS there is a literature relating efficiency to time, technique (a concept broader than technology), cooperation, competition, justice, and many other matters. The year 2009 has already produced a bumper crop of studies from diverse theoretical perspectives and from these I have selected four for review. They present four very different approaches to understanding efficiency, but provide complementary rather than conflicting treatments of the topic. Rather than reviewing each of these books in terms of their own disciplinary orientations – management, ergonomics/reliability engineering, law, and environmental economics – I propose to review them with an eye on the LIS discourse of efficiency: just what is meant by “efficiency” in LIS? Is efficiency a technical issue? An ethical issue? A necessity? Economic sense or nonsense? Managerial propaganda or just a word we use to justify organizational changes? These are precisely the kinds of questions raised and discussed in the books reviewed here, and in fact the meaning of the word “efficiency” in managerial discourse is the main topic of Callender’s Efficiency and Management.

Callender remarks in his preface that his experiences in the business world and government led him to try “to better understand the management use, even worship, of the term efficiency” (p. xiii). The method he chose was to examine the use and varied meanings of the term efficiency in the literatures of engineering, management, economics, accounting, human resources management (HRM) and the procurement profession, adding case studies of economic commentary in the mass media, and of the railroad and procurement industries. What he argues is “the demise rather than the rise of efficiency” (Callender, p. xiii), suggesting that unlike the industrialized nations of the west, the “rising economies in Asia and other parts of the world … seem to possess a concept of how to enact efficient practice” (Callender, p. xiii). His principal objective is, he writes:

• •

  … to demonstrate that, despite the historical management support for the notion of technical efficiency until the middle of the twentieth century, the notion of efficiency now has a limited impact on contemporary management practice despite its constant usage in management discourse. (Callender, p. 3)

and what he finds confirms:

• •

  … that the notion of descriptive efficiency has been developed as a populist concept in managerial discourse and is typically interpreted in simplistic, normative terms and thus has limited technical meaning in management practice. In the hands of plausible management commentators who have seemingly assumed that management efficiency will emerge from the adoption of their various prescriptions, the term provides status without necessarily creating substance (Callender, p. 3).

What his study revealed and what he is arguing is not an objection to efficiency but to the meanings that the term has come to have in the fields investigated and the way the term is used. Efficiency has become, he argues, “an ideological statement of support for any management intention, rather than a practical means to inform a range of management actions” (Callender, p. 3). As a wonderfully humorous example of this he remarks that:

• •

  … it is unlikely that Chief Executive Officers (CEOs) would wish their remuneration to be linked to least cost concepts even though their rhetoric concerning corporate efficiency and shareholder Return on Investment may often promote such views (Callender, p. 19)

Defining efficiency in terms of least cost is an example of what he calls “economic efficiency” (abbreviated in the book as EEec) which he distinguishes from other meanings of efficiency such as descriptive efficiency, allocative efficiency, technical efficiency and dynamic efficiency. (The multiplicity of meanings and the terms he provides for them are confusing enough; his resort to abbreviations makes reading very difficult.) The problem with understanding efficiency in terms of an economic definition like “least cost” or “greatest benefits relative to cost” is that these terms:

• •

  … may be quantified when applied to a single situation, but both terms suffer the legacy of the descriptors least and greatest, both of which are subject to individual interpretation. How does an organization know that it has achieved its least cost goals? Can the term least ever imply an absolute value, or is it subject to change over time? What are the measurable greatest benefits relative to costs for an organization? Are these values ever finite or will greatest benefit vary as the nature and value of costs change over time? (Callender, p. 20).

Callender traces the history of the term “efficiency” in the management literature, revealing a move away from the use of efficiency to mean “technical efficiency” (maximizing performance) towards “economic efficiency” (least cost, etc.), a shift away from the engineering perspective focused on what one intends to achieve, to the managerial perspective narrowly concerned with cost cutting and profit. His case study of the railroad reforms of the 1990s in the UK and New South Wales is a narrative of the triumph of economic efficiency over technical efficiency among those responsible for the rail system “in an environment where system design and operation requires the maintenance of quite rigid professional operating standards” (Callender, p. 146-147). The new policies adopted were service-based models that sought “to improve general levels of customer service, reduce the cost to the state of operating the public rail network and increase the efficiency of labour” (Callender, p. 146). The officers of the State Rail Authority of New South Wales were committed to “the competitive outcomes-focused ideology of Economic Rationalism” and assumed that “better service was the path required to rejuvenate rail, with safety and operating efficiency being somehow looked after by the myriad of [outsourced—DB] contractors that now undertake rail maintenance.” These new managers, he writes, “seemingly believed that a laissez-faire approach to technical efficiency (TEmp) would not have any significant consequences” (Callender, p. 168).

Within a decade of the railroad reorganization in New South Wales judicial inquiries into accidents led to recommendations to reverse most of the policy changes of 1996. Callender writes that the primary achievements of those policies – made in the interest of achieving greater efficiencies – have been an intense media scrutiny of the practical results of policies, a judicial review of an accident that “linked a decline in safety standards in the interest of on-time running”, a major reduction in staff, “a limited improvement in financial performance” and new regulatory bodies that increased operation costs (Callender, p. 167). And although the New South Wales system is almost back to where it was before, the rail system has “yet to deliver either technical efficiency (TEmp) by demonstrating higher output from existing resources or … by delivering rail owners least cost outcomes” (Callender, p. 171). The government promoted efficiency and a competitive structure “without regard to the history, culture and technical consequences of such a change” (Callender, p. 168). The price of this “efficiency adventure” Callender writes, “of lives lost, careers destroyed or damaged, and political credibility and money will never be known” (Callender, p. 169).

Callender’s final chapter opens with a concise summary of the results of his study:

• •

  The evidence gathered in this study suggests that the management discipline has drifted into allowing the notion of efficiency to be typically defined within the confines of the economics discipline as any activity that results in a lower or least-cost outcome, whether or not this outcome can be demonstrated. If this is the case, then the evidence also implies that efficiency has come to be seen as an ideological expression rather than one that defines standards of management or organizational performance. In short, the word efficiency has become a meaningless figure of speech in the vocabulary of managers. The exact meaning implied by the management user is almost unimportant, as the mere use of the term suggests that the outcomes being reported are justifiable. (Callender, p. 182)

One need only look into the literature of library management to verify Callender’s conclusions for LIS. Yet what of those of us who really do sense the need to discuss matters related to efficiency, those for whom it is not an ideological weapon or propaganda? For such readers Erik Hollnagel’s book is a perfect complement to Callender’s.

The first thing to note about The ETTO Principle is that unlike Callender’s book it is a breeze to read. Hollnagel writes well and writes for “people who, for one reason or another, are reluctant to start on a conventional textbook or work of science” (Hollnagel, p. iv). Rather than offering the reader a dense, terminologically complicated text filled with references and scholarly apparatus, he has chosen to write for the lay reader and to conclude each chapter with some notes on the literature, there for your interest, to be pursued or skipped over at will.

The second thing to note is that Hollnagel writes from the perspective of one of the world’s leading researchers on why things sometimes go wrong (see the subtitle of the book) in the operation of technical systems. His definition of efficiency is neither ideological nor economic but practical. The first discussion of efficiency appears in his remarks on the Stop Rule:

• •

  Since the purpose of an accident investigation is to find an adequate explanation for what has happened, the analysis should clearly be as detailed as possible. This means that it should not stop at the first cause it finds, but continue to look for alternative explanations and possible contributing conditions, until no reasonable doubt about the correctness of the outcome remains …

  Ending an analysis when a sufficiently good explanation has been found, or using whatever has been found as an explanation when time or resources run out, even knowing that it could have been continued in principle, corresponds to a criterion of efficiency.

  We shall call the first alternative thoroughness, for reasons that are rather obvious. And we shall call the second alternative efficiency, because it produces the desired effect with a minimum of time, expense, effort or waste. (Hollnagel, p. 12-13).

That is a pretty simple definition and one upon which Hollnagel will expand a few pages later where he also offers an expanded definition of “thoroughness”.

The third thing to note is already in the passage just quoted: efficiency is defined in relation to its opposite. The main point of the book is that efficiency and thoroughness are “both necessary but rarely simultaneously possible”. This is what the ETTO Principle states:

• •

  In their daily activities, at work or at leisure, people routinely make a choice between being effective and being thorough, since it rarely is possible to be both at the same time. If demands to productivity or performance are high, thoroughness is reduced until the productivity goals are met. If demands to safety are high, efficiency is reduced until the safety goals are met (Hollnagel, p. 15)

Hollnagel then expands on the definition of efficiency:

• •

  Efficiency means that the level of investment or amount of resources used or needed to achieve a stated goal or objective are kept as low as possible. The resources may be expressed in terms of time, materials, money, psychological effort (workload), physical effort (fatigue), manpower (number of people), etc. The appropriate level or amount is determined by the subjective evaluation of what is sufficient to achieve the goal, i.e. good enough to be acceptable by whatever stop rule is applied as well as by external requirements and demands (Hollnagel, p. 16-17).

Efficiency is necessary because time and resources are both limited; thoroughness is necessary “to make sure that we do things in the right way, so that we can achieve what we intend, and to avoid adverse consequences” (Hollnagel, p. 17) these latter being major producers of inefficiencies. Hollnagel’s emphasis on time is particularly interesting; Callender also noted that demands for efficiency often went hand in hand with a failure to consider temporal developments and potential consequences, and Merlini (2009) puts our changing experience of time as a result of technological developments at the centre of his analysis. Classical economic decision theory, Hollnagel reminds us, assumes “that alternatives as well as criteria are constant while the decision is made, that hence time does not exist” (Hollnagel, p. 25). In the real world, however, when one wishes to do something:

• •

  … there are always two options. One is to wait, to gather more information, to see how things develop, or just to hope for a greater level of certainty – or less uncertainty. The other is to go ahead on the assumption that the situation is known well enough and the alternatives are clear enough... It is this dilemma between time to think and time to do that is at the heart of the ETTO principle. (Hollnagel, p. 28)

• •

  The ETTO principle names a phenomenon or a strong characteristic of individual – and collective – performance, namely that people in dynamically changing, hence unstable and partly unpredictable situations, know that it is more important to do something before time is up, however imperfect it may be, than to find the perfect response when it is too late (Hollnagel, p. 58)

This will remind many readers of the discussions in the library literature about rapid versus full cataloging, timely access versus “the perfect record” and so on. This is exactly Hollnagel’s terrain and although librarianship does not enter the discussion, he does discuss the time sensitive data collection, analysis and dissemination work of the Canadian office of Aeronautical Information, Regulation and Control. This office is required “to produce a weekly report that is as accurate as possible” (Hollnagel, p. 73) – and it succeeds.

Of particular interest for library operations is Hollnagel’s discussion of “collaborative ETTO” in chapter six. All actions and especially collaborative actions do not occur in isolation but follow from and influence the actions of others. We can think of our actions in the context of the actions of others either from the viewpoint of efficiency or of thoroughness:

• •

  In such cases, thoroughness means that the person does not simply accept the input he or she receives from somewhere or from someone else, but instead makes an effort to confirm that it is correct … Similarly, thoroughness would mean that the person considers the possible side-effects and secondary outcomes of what he or she produces as output, in a sense adopting the mindset of whoever is going to work on the results. Similarly, efficiency means that the person trusts that the input he or she receives is correct, i.e. that the previous person was thorough. Efficiency also means that the person assumes that the next person, whoever is going to work on the results, will make the necessary checks and verifications, i.e. that the next person is thorough (Hollnagel, p. 119).

Since there is never enough time or resources to check everything at each step of the way we have to trust what others have done, i.e. the choice for efficiency is forced upon us:

• •

  It is as if everyone, including ourselves, reasons in the following way: “I can allow myself to be efficient, because the others will be thorough.” If only some people do that, the system may be able to correct itself and to find a balance of functioning that is both reasonably effective and reasonably thorough. But if everyone begins to work in this way, for instance because of systemic pressures, the net result may be that something goes wrong (Hollnagel, p. 120).

This describes the world of cooperative cataloging very well: its appeal, its necessity, its potential catastrophe. We are eager to enter into cooperative agreements and automate, and outsourcery has become our new religion because it means we will no longer have to perform the required labor nor deal with a staff but only pay the lowest bidder. And we sit and wait for someone else to do our work for us (cooperation), or we significantly limit our knowledge and control over the results (outsourcing and automation). How do we avoid this scenario in which we each expect everyone else to do what we do not have the time or resources to do? From the analysis of case studies of accidents, near accidents and successes Hollnagel concludes that socio-technical systems work only when and only because everyone makes “approximate adjustments to their work” (Hollnagel, p. 123). We must understand, he argues:

• •

  … that what anyone does depends on what others have done, what others do, and what others are going to do – and that everyone is in the same situation. In order to be able to improve safety we must acknowledge that the most important part of the – dynamic and unpredictable – environment is what other people do. It is the ability of people to adjust their performance in response to the adjustments that others have made, make, or are going to make, that makes the social system strong (Hollnagel, pp. 123-124).

Technical systems are not like social systems, and socio-technical systems are not like either social or technical systems considered separately. If I may put my own interpretation forward, I would suggest that, following Ellul (1964), technical systems embody the search for efficiency, while social systems embody the search for inefficiency, i.e. freedom, for it is only in the space where anything can happen that the new and the unforeseen – the future – become possible. Any socio-technical system must seek a balance if we are to avoid the unbearable alternatives of a future that is totally controlled or totally out of control.

The ETTO principle describes that search for a balance which, because of our temporal and corporeal existence, is always a trade-off. Because both efficiency and thoroughness are necessary, we are left with a paradox:

• •

  … efficiency in the present presupposes thoroughness in the past, which paradoxically means that thoroughness in the present is necessary for efficiency in the future (Hollnagel, p. 149).

And that leads us back to Callender’s question “Are these values ever finite or will greatest benefit vary as the nature and value of costs change over time?”

The recent literature discussed by Callender, like the library literature, views efficiency from a very narrow perspective: cost-cutting. The beauty of Hollnagel’s book is that he looks at efficiency from the perspective of temporal processes within socio-technical systems, recognizing their dynamics and unpredictability in light of the full range of human interests, needs and susceptibilities. Mathis’ Efficiency Instead of Justice? provides yet another complementary perspective, examining the meaning of efficiency in light of legal and economic justice.

Part One of Efficiency instead of Justice? is devoted to understanding the economic analysis of law, in particular the notion of efficiency in welfare theory. Part Two discusses the philosophical foundations of the economic analysis of law, dealing with Adam Smith, Jeremy Bentham and John Rawls in order. In Part Three we move to the contemporary debate surrounding Posner’s ideas, the author’s critiques and his conclusion. The book is particularly relevant to recent debates in librarianship not only because of its discussion of Homo economicus, but also because of its discussions of welfare economics and distributive justice, matters of significance to librarians dedicated to access to information and worried about the digital divide. There is also his critique of Bentham, whose utilitarianism was recently resurrected as a philosophy for librarianship precisely in the context of developing more efficient library operations (Banush and LeBlanc, 2007).

The second chapter, “Efficiency criteria” presents two economic understandings of efficiency – Pareto efficiency and Kaldor-Hicks criterion –and criticisms of them. The criterion for Pareto efficiency is “any change that puts one member of a society in a better position without making somebody else worse off” (p. 33). The Kaldor-Hicks criterion states that “a change is an improvement … if the gainers value their gains more highly than the losers their losses” (p. 39). I will not dwell on these theories nor on Mathis’ criticisms, but one objection he makes to adopting a consistent policy following the Kaldor-Hicks criterion is that humans “need to have just treatment meted out – in specific cases and, most importantly, case by case! – to themselves and, indeed, to others” (quoted from the English edition, p. 49). This observation parallels my own understanding of the nature of bibliographic description (it is an activity of attention to the facts case by case), and is pertinent to discussions of automated systems of cataloging as much as the administration of justice. Summarizing the chapter Mathis writes that both Pareto’s and Kaldor-Hicks’ understanding of efficiency in economics:

• •

  … contain not only weak, uncontested value judgements but also strong ones. Far from making analysis of normative questions superfluous, the demand for efficiency therefore has the opposite effect: it brings into sharp focus the need for closer scrutiny of the relationship between efficiency and justice. (p. 49)

That relationship he treats directly in Chapter Nine, to which we now turn.

Mathis first looks at the relationship between efficiency and justice in Hollnagel’s terms as a “tradeoff.” He rejects the economists’ assumption that this relationship is one of conflict, assuming instead that “justice and efficiency are substitutable, up to a certain point” (p. 185) and proceeds to discuss them in terms of a trade-off on the values-level – “how much justice a person or a society is prepared to sacrifice in order to achieve more efficiency (or vice versa)” (p. 185-186), and a production trade-off: “how much justice must be sacrificed in order to achieve a certain level of efficiency (or vice versa)” (p. 186). For these trade-offs we can substitute “thoroughness” for justice and get Hollnagel’s ETTO, or “service to the users” and we get the debate in the library world. And here Mathis returns to the concept of efficiency, and asks the question “Is efficiency a goal at all?”:

• •

  Efficiency is far rather an instrument for achieving other social goals. According to Dworkin, a trade-off between means and ends makes no sense – unless efficiency is thought of as a “false target” for other goals … Therefore efficiency is only of limited use as a “false target” for increasing social utility in the utilitarian sense (p. 191).

Certain policies in the library world suggest that efficiency has indeed become a false target rather than a means to some other social goal, and as Mathis/Dworkin has argued, this makes no sense. In another recent and fascinating book (that perhaps I should also have included in this review) Merlini has argued that this “meaningless efficiency [efficienza insignificante]” is “putting the priority of the answer ahead of the question” (Merlini, 2009, p. 66) a way of stating the issue that I find particularly revealing. Ellul took a related approach to technical development, arguing that the chief aim of technique was efficiency, and that in the technical approach to any problem “The multiplicity of means is reduced to one: the most efficient” (Ellul, 1964, p. 21). Making efficiency the aim towards which we strive and determining the means towards that end solely according to criteria of efficiency leads to absurd decisions and just those conflicts that Mathis insists are not necessary. In the end we get injustice instead of efficiency, and, as Morel (2002) and Kerdellant (2000) have documented, business failures and technological disasters. None of these would any of us consider to be efficient.

Mathis does not leave us with a trade-off, however, since he argues that this is a one-dimensional approach. Instead, he argues, rather than a monocausal relationship “there are multiple interdependencies between the two goals, and while they conflict in some respects, in many other ways they actually stand in a harmonious or at least a neutral relationship to one another” (p. 198). And in his conclusion he argues that “the endeavour to realize both goals need not always be a competitive trade-off, and can in fact be undertaken cooperatively to a large extent. Piana (2008 – yet another recent book that I could have included in this review) argues similarly in the context of globalization, insisting that “economic rationality and ethical rationality do not go hand-in-hand any more than they are in opposition, rather they converge in a common terrain” (p. 10). Piana in fact takes as his premise Mathis’ conclusion and proceeds to discuss that cooperation in terms of the ethics of solidarity in a global economy in which information technologies “are no longer inscribed in the order of means, but are more and more acquiring the character of ends” (Piana, 2008, p. 5). The focus on cooperation and the recognition that efficiency as a means is not necessarily in opposition to the social ends which are our goals is all good news – and matters apparently not well understood among librarians – but we have not yet dealt with the Jevons Paradox.

The Jevons Paradox was stated succinctly by Jevon’s as follows:

• •

  It is wholly a confusion of ideas to suppose that the economical use of fuel is equivalent to a diminished consumption. The very contrary is the truth. (Jevons, 1866, p. 123, quoted in Polimeni et al. (2009), p. ix)

In his introduction to Polimeni et al. (2009) Tainter spells out the consequences of this paradox:

• •

  It suggests that efficiency, conservation and technological improvement, the very things urged by those concerned for future energy supplies, may actually worsen our energy prospects. (Tainter, in Polimeni et al., p. x)

The Myth of Resource Efficiency has, apart from the introduction and conclusion, only three chapters. The first of these is a review of the classical and contemporary economics literature on the matter mentioned by Jevons in the passage quoted above. In the first chapter (by Blake Alcott) the question is put directly in the author’s introduction: “is efficiency part of the solution or part of the problem?” (p. 13). Alcott follows this lead question with a discussion of the definition of efficiency according to the economists discussed:

• •

  Throughout the following examination of our author’s definition of efficiency it is axiomatic that efficiency denotes a ratio. The numerator is output and the denominator is (energy) input. “Efficacy”, or “effectiveness” or, more ambiguously, “power” denote in contrast the causation of a given amount of output regardless of cost or input. Ontologically, the thing that is efficient is the input … We are not investigating consumption efficiency – for example boiling only the amount of water needed for the cup of coffee. (p. 13)

He then makes the interesting observation that it matters how we define an increase in efficiency, for understanding an increase of efficiency as “less input per unit of output … biases our thinking by holding output constant and looking at what could be saved” whereas understanding the same process in terms of “more output per unit of input … biases it by highlighting increased output with perhaps no saving” (p. 16). A good portion of this chapter is devoted to the literature on “rebound” and “backfire”. Rebound is when an increase in efficiency causes or contributes to an increase in resource use. Fuel efficient cars lead to (whether contribute to or cause is a key part of the debate) people driving more. When the gains in efficiency lead to so much more driving that the increase in fuel consumption is more than 100 per cent of that “saved” by the efficiency measure, this is called “backfire”. Although the authors do not make the connection explicitly, it seems that backfire is a phenomena associated with thinking of efficiency in terms of “more output per unit of input”, much like we read in the library literature of “more, better, cheaper, faster.”

The first chapter was fascinating to read, much more enjoyable than Callender’s discussion of efficiency in the management literature (not the fault of Callender, but of the management literature). But as one might expect in a review of scholarly literature, scholars disagree. There are no answers, and those looking for answers rather than a good question (such as that posed by Alcott and quoted above) will have to pick up a less interesting book. That would however be a shame for the second chapter is remarkably appropriate for anyone working with information technologies. That chapter, by Giampietro and Mayumi, concerns the evolution of complex adaptive systems, both biological and socio-technical.

The authors begin with a statement of three conceptual problems for the study of the relationship between efficiency improvements and rebound effects. The first is how to define and measure efficiency, the second is how to distinguish efficiency improvements “due to a change in technological coefficients (when the system performs ‘the same set of transformations’; but ‘better’)” from those improvements due to changing the nature of the task, “when the system finds more convenient methods to perform ‘something else’ instead of the original set of transformations” (p. 80). It was this second conceptual problem that particularly caught my attention because this is precisely the problem we are facing when we try to understand how to use and evaluate the new technologies and techniques available today. For example, is the automatic generation and exchange of metadata really more efficient than human created metadata, or are these two very different tasks producing two very different (though perhaps superficially similar) products which cannot be compared in terms of efficiency precisely because of the differences? If we choose the “more efficient” system we are in fact choosing a very different system and changing entirely the “complex adaptive system” that our current information infrastructure most certainly is. This is in fact the heart of Giampietro and Mayumi’s topic, the purpose of whose chapter “is to provide a different perspective on the discussion about the link between increases in efficiency and sustainability” (p. 80) and it is a truly fascinating and provocative discussion.

“Living systems when evolving in time have the peculiar ability to ‘become something else’” the authors remark and then extend the applicability of the statement to socio-technical systems with a discussion of technological change in agriculture. Technical innovation and efficiencies have eliminated animal power from agriculture but this, they note “implies that improvements in efficiency within a given context (in this case farming in the oil era) do imply a reduction of adaptability in the long term (if we will run out of oil)” (p. 80). They anticipate one of their conclusions in the comment “These theoretical issues imply, when dealing with evolutionary trajectories, that it is impossible to use the concept of efficiency for planning the best course of action” (p. 81).

At this point the reader may recall that both Callender and Hollnagel stressed the importance of time and temporal developments for understanding efficiency and the effects of efficiency measures. An ecological perspective cannot be narrowly bounded, whether temporally or geographically, but this kind of forward glance is not at all like the futurology and crystal ball gazing one finds among prognosticators of technological invention and library futures. If only the library world had prognosticators with the perspective of Polimeni et al.!

The critique of formal models for evaluating and planning for efficiency is a central element in this second chapter and provides an argument from evolutionary theory (and in English) to complement and strengthen Górny’s (2008) argument (in Polish) on evaluating efficiency in libraries. The more useful that formal models are “for increasing improvements in efficiency, the quicker the status quo will change and the more likely it is that these formal models will become useless for making tong-term predictions” (pp. 90-91). Continuing their discussion of efficiency in relation to temporal developments the authors note that “alternative (and also contrasting) assessments of efficiency can be found when considering simultaneously tasks referring to different temporal horizons” (p. 95). And those alternative and contrasting perspectives “are necessary to preserve diversity”, for:

• •

  … a successful surviving trajectory of evolution must result in the ability to establish an impredicative loop between increases in efficiency, an attribute very relevant in the short run, and an increase in adaptability, an attribute very relevant in the long run. If only one of the two strategies is adopted – increasing adaptability by reducing efficiency or increasing efficiency by reducing adaptability – a negative side-effect will show up either in the short or the long term (p. 97).

The authors elaborate on this matter and the farming example above later on in terms that are strikingly reminiscent of the many calls for change in the library, yet at the same time just as strikingly different. They define adaptability as:

• •

  … the ability to adjust our own identity in order to retain fitness in the face of changing goals and changing constraints. Fitness means the ability to maintain congruence among a set of goals, the set of processes required to achieve them and constraints imposed by boundary conditions … Therefore adaptability requires the ability to preserve diversity (an adequate option space) in terms of both possible behaviours and organizational structures (p. 122).

When efficiency is “defined at a particular point in space in time, according to a particular interpretation of this term” – which it must be for practical action – then “in order to increase the efficiency … we have to eliminate” less efficient activities and support the efficient ones – from the perspective of that temporally and spatially limited understanding of efficiency. And here they go back to the farm:

• •

  Driven by technological innovations tailored on this interpretation of efficiency – such as the green revolution or genetically modified organisms – agricultural production all over our planet is converging on a very small set of standard solutions (commercial seeds, technological packages, and economic demand heavily affected by transnational corporations and globalized markets). On the other hand, the “obsolete” agricultural systems of production, those that are being abandoned all over the planet, may show a very high performance if a different set of goals and criteria – rural employment, ecological compatibility and preservation of biodiversity – were adopted (p. 122).

In other words, the question that remains is the one I asked at the beginning of this review: efficient at doing what? It is here that efficiency becomes a political OR an ideological matter (and sometimes both):

• •

  Any given perception/representation of the external world based on a particular formal model must necessarily reflect a set of choices made by a special story-teller about the selection of a relevant narrative for a given state of affairs in relation to a given set of goals … As Schumpeter aptly remarked, “[a]nalytical work begins with material provided by our vision of things, and this vision is ideological almost by definition” (p. 129)

The final chapter by Polimeni provides an empirical macroeconomic analysis of rebound and backfire across a wide range of countries and resources. Although the whole book is devoted to the Jevons paradox in reference to energy resources, the authors stress that the paradox appears to apply to the use of any and all resources, including, presumably, information resources. Polimeni’s general statement appears to describe exactly what librarians and other information professionals have experienced over the past decades:

As a resource becomes more efficient to use, and more affordable, current technology will be used more or new technology will be introduced that contains more options and features. (p. 147)

Now, however disconcerting this may seem in relation to oil and other non-renewable resources, the LIS profession has been dependent upon this “paradox” and in fact many information professionals are delighting in it and dreaming of a glorious future in which we will be the crown jewels of the information society. If improvements in efficiency are counter-productive and actually promote consumption, then the labour required in the future will increase rather than decrease. We immediately think job security, but perhaps there is more to it than that. What does this mean for the development of efficient search and research, web-authoring, publication and scholarly communication of any kind? In spite of my own interest in the matter of policies and organizational structures and scholarly communication in libraries, this book made me think that the arguments presented are probably more important for those whose interest is in the design of technical systems such as search engines. In the conclusion the authors mention Zhu Yuan-Chang’s chessboard, a story illustrating a hypercycle. They remark:

• •

  Hypercycles, or positive autocatalytic loops, when operating without a coupled process of control (and damping), do not survive for long; they just blow up (p. 173).

I really wish the authors had discussed the production and use of information resources in relation to the Jevons paradox, but perhaps this review will inspire some reader to undertake the matter. If the Jevons paradox pertains to information resources like all other types of resources, then does their warning that relying on “efficiency and technology as a solution is foolhardy” (p. 3) apply to LIS as well? For those not afraid to think outside the LIS box, this book is a challenge indeed.

Callender, Hollnagel, Mathis, Polimeni et al. – what has all this to do with the design and use of information technologies and the work that librarians do? For one thing it should make us see the insufficiency and even misleading nature of the greater part of the discourse on efficiency that we find in the LIS literature. We appear to have made little or no progress since Robinson (1920) stated “The measure of the efficiency of any library must be the measure of its usefulness, all else being plant and machinery and operators contributing to that end,” for recent writers on efficiency treat the topic as an issue separate from usefulness or effectiveness. Chen (1997) stated “efficiency refers to resource utilization efficiency, rather than an evaluation of effectiveness” and Shim (2000) similarly distinguished effectiveness (usefulness for the users) from efficiency (technical efficiency, or the ratio of outputs to inputs) and assumed that the latter can be evaluated separately from the former.

During the past few years innumerable references to efficiency have appeared in the literature of library management, system design (searching efficiency) and discussions of library legislation, yet a search of these literatures revealed that with few exceptions (such as Hernon and McClure, 1990) efficiency is assumed to be an unproblematic concept. Efficiency in the LIS literature is rarely defined, and when a definition is given, it is often reductionist and simplistic, such as Vitaliano’s definition:

• •

  Efficiency is defined as whether or not a library could reduce the inputs it uses equiproportionately and still produce the same output. Inputs are defined programmatically: holdings, opening hours, serials and new books. Output is internal and external circulation. (Vitaliano, 1998, p. 107)

Hammond (2009) in an otherwise interesting article writes of “efficiency” without any qualification throughout the paper, but also refers to technical efficiency – i.e. production efficiency, the object of his study – allocative efficiency and economic efficiency. Unfortunately he nowhere mentions effectiveness. (His article is a good example of the multiple terms and related problems noted by Callender.) He does however recognize the inadequacy of earlier studies, noting that “Efficiency is, as a result, a potentially more complex phenomenon than many earlier studies have implied” (p. 3).

With effectiveness eliminated from the discussion, many authors proceed to evaluate efficiency without any relation to other factors that determine the success of people’s choices and actions (e.g. time, cooperation, competition, communication, learning). Efficiency as a value both unquestioned and in no relation to any other values appears in Barkut (2008) who was content to insist that the increased use of IT needs to result in a more efficient operation. Indeed many references to efficiency limit the discussion to the efficiency of the technologies in doing what they are designed to do, not the broader context of people in their various tasks and what they need to do. We find an example of this kind of treatment in Mansourian (2008). The author proposed a new conceptual measure (web search efficacy, or WSE) to evaluate the performance of web searches, but as WSE is based on user expectations of “quick and easy” it is really a measure of user perceptions of how quick and easy searching is – which certainly has something to do with efficiency but is not the same thing as asking the users about whether the system was effective. Fischer and Schwan (2008) went a bit further, treating efficiency as synonymous with speed – the faster the results are obtained, the more efficient the system – arguing that what Microsoft engineers thought would make for more efficient use of their products (adaptively shortened pull down menus) actually resulted in an increase of inefficiency, i.e. a decrease in speed of obtaining results. Questions of ease and effectiveness in obtaining the desired results did not enter into their evaluation.

In one of the best articles on the topic Górny (2008) asked why our assessments of library efficiency have been of little use and suggested that our efforts have been devoted largely to methodologies and techniques, when our focus ought to be on library objectives. As mentioned above, Górny’s remarks on the effect of rapid technological change on evaluation are similar to the criticism made in Polimeni et al. where we read “when trying to predict changes associated with evolution, any formal model is bound to become obsolete” (Polimeni et al., p. 110). Lazcano Herrera and Font Graupera (2008) criticize what they consider the misplaced emphasis on technologies and system performance, suggesting that we need to focus on library users’ experiences with technologies, looking not at what technologies are designed to do (however efficiently) but at whether library users can actually use them to do what they want or need to do. When efficacy is understood to be central to the notion of efficiency, then to understand efficiency in libraries the first and most important task is to understand the goals of library users, goals that are largely related to social and normative practices. When we have made clear to ourselves why libraries exist, for whom and to serve what purposes and practices, then and only then may we determine what means (technical efficiency) will best serve the library’s users, and how our policies (economic and allocative efficiency) will effect our service over time (dynamic efficiency).

The political and ideological character of the discourse on efficiency was noted not only by Callender for whom it was the central issue, but also by Mathis and Polimeni et al. The 2008 report of the Library of Congress Working Group on the Future of Bibliographic Control presents an example of that ideological discourse. We read therein that “Libraries must work in the most efficient and cooperative manner to minimize where possible the costs of bibliographic control,” and throughout the document efficiency is equivalent to a reduction of cost or cooperation (the managerial euphemism for requiring that someone else do the work). It appears that the same ideology was at work in the drafting and passage of the German and Belgian legislation mentioned in BuB: Forum Bibliothek und Information (2008) and Cauwenbergh and Lekens (2009).

The advocacy of cooperation noted in the preceding paragraph points us to the most important issue of all for working with global technical systems: efficiency in relation to cooperative action, a topic that Hollnagel has discussed in more detail than the other authors. Some librarians have made a serious mistake in assuming that joint exploitation of a technical system is the same thing as cooperation. Indeed it is noteworthy that demands for efficiency are frequently coupled with an extravagant praise of cooperation – e.g. in Cauwenbergh and Lekens (2009) and the Library of Congress report mentioned above – perhaps because cooperation is misunderstood as being nothing more than letting someone else do the work – outsourcing under another name?

Sixty years ago Roethlisberger (1949) complained that the “striking contrast between technical efficiency on the one hand and matters of human cooperation on the other presents the number one problem of our present industrial civilization” (p. 233). His comments on efficiency and the relationship of technology to social structure are remarkably appropriate to today’s “information society” as well:

• •

  I find little justification for the prevailing assumption that so long as we turn out goods efficiently of good quality and of low cost, matters of cooperation can be left to chance. I find little evidence for the popular beliefs that cooperation is a matter of logical and technical contrivance or a matter of verbal exhortation – something that can be willed into being by verbal persuasion or efforts of personality. I find that there are just as brute and stubborn facts that determine matters of cooperation as there are brute and stubborn facts that determine matters of technical efficiency … (Roethlisberger, 1949, p. 233)

• •

  Many of the changes modern technology originates can collide head on with the social organization of the company and its attempt to maintain internal stability—a necessary precondition, as we have seen for cooperative behavior. With the very best of intentions, modern technology can unwittingly foster the segmentation of the social structure of industry into groups with radically different points of view. It can unwittingly assist in the development of rigidities of relationship between segments of the structure that make cooperation difficult, if not in some cases impossible. The patterns of behavior produced by modern technology do not in and by themselves make for cooperation (Roethlisberger, 1949, p. 237)

We continue to think of our use of technical systems as though the system itself enables not only the performance of the task but cooperation as well, as though our use of the system has no consequences for others elsewhere or in the future, as though efficiency is a purely local and private matter for us here and now, as though we live isolated within a world which we can exploit as we please rather than in a world in which all of our actions affect everyone everywhere else, including ALL future generations. This is a mistake with serious consequences, and Hollnagel’s discussion of cooperation in socio-technical systems, Piana (2008) on efficiency in the context of globalization, and Polimeni et al. from their macroeconomic and ecological perspective have all provided evidence and arguments that help us understand efficiency, cooperation (and exploitation) in the context of shared and finite resources.

What the books reviewed above can do for us – if we read them carefully and together – is to lead us to see that the appeal to efficiency is not a simple matter at all, and that when efficiency becomes an end rather than a means we can easily get distracted from what it is we are actually trying to accomplish, and perhaps prevented from accomplishing it. We need to think ecologically, globally and ethically about efficiency (and economy, cooperation and technologies) in libraries and how our actions today are going to shape the world for our children and their children. For if Hollnagel is right that efficiencies tomorrow will require thoroughness today, our children are going to hate us for sure.

David BadeUniversity of Chicago