Operational Risk, Translation, and Globalization

This paper compares a translation of a global (more specifically, European) regulation into two local contexts, setting this process in a broader context of the all-pervading risk management. The two countries are Sweden and Poland, both relatively untouched by the current financial crisis, and the regulation is Basel II Accord. In both countries, the translation is shaped by the past history, and the present circumstances. The results show that, in spite of local differences, there is a common belief in quantification of risks as the main remedy and therefore the main way of managing them. Abstract and vague formulations, combined with sophisticated calculation techniques, win over the complications of actual practices. The role of researchers in this process is also examined. A study illustrates also the advantages of translation theory versus diffusion theory of spreading of ideas.

This paper compares a translation of a global (more specifically, European) regulation into two local contexts, setting this process in a broader context of the all-pervading risk management. The two countries are Sweden and Poland, both relatively untouched by the current financial crisis, and the regulation is Basel II Accord. In both countries, the translation is shaped by the past history, and the present circumstances. The results show that, in spite of local differences, there is a common belief in quantification of risks as the main remedy and therefore the main way of managing them. Abstract and vague formulations, combined with sophisticated calculation techniques, win over the complications of actual practices. The role of researchers in this process is also examined. A study illustrates also the advantages of translation theory versus diffusion theory of spreading of ideas.
In 1956, in a book entitled Bourgeois Morality (English version published in 1986), Polish sociologist Maria Ossowska (1896Ossowska ( -1974 described the historical development of a bourgeois morality. The description ended in contemporary Poland, a newly socialist country, but already past the horrors of Stalin's regime. One of her last chapters included the analysis of a brochure issued by the Polish Saving Bank (PKO) in 1948. A product of the new socialist regime, it revealed to Ossowska's critical eye strong sediments of the capitalist past ("For the sociologist, it is intriguing to see the detritus of the past continuing by force of inertia into the present", [Ossowska, 1986, p. 123]). The title of the brochure, "Industry and frugality", alluded -probably inadvertently -to Benjamin's Franklin's maxim "Be industrious and frugal, and you will be rich. " The text of the brochure continued in the same spirit, and ended with what appeared to be an unintended allusion to a version of Aesop's fable about an ant and a grasshopper by Russian fabulist Ivan Krylov (1769-1844. In the fable, the ant was working and saving all summer long, while the grasshopper spent the summer singing, and now wanted to borrow money from the ant. "Go and dance then", responded the ant. The Polish Saving Bank said almost the same thing, in verse form. Having compared the two texts, Ossowska (1986) hastened to assure readers that:

Operational Risk, Translation, and Globalization
Operational Risk, Translation, and Globalization The texts say more and less than that what their authors' wish; translations (linguistic or not) can change the text beyond recognition (for better and for worse), and institutional sediments are much more resistant to change than the eager change agents wish. In the global world, ideas travel around the planet, but are then locally translated. The result may be that the same idea differs every place it lands; that different ideas may lead to similar practices; and that the final combination of global ideas and local practices is almost inevitably difficult to foresee, but fascinating to study.
In what follows, I address the concept of operational risk management as launched by Basel II Accord, and the way it has been translated in two countries -Sweden and Poland -both relatively untouched by the recent (and continuing) financial crisis. Another case of a translation of risk management completes the analysis, which is preceded by a short introduction to the sociology of translation -the framework used in this text.

Some notions from the sociology of translation
In order to denote the transfer of ideas and practices from one context to another, scholars have traditionally spoken of diffusion, following Rogers (1962), who borrowed this concept from Gabriel Tarde (1890Tarde ( /1903).
But to most people "diffusion" suggests a movement subject to the laws of physics, quite contrary to Tarde's intentions; he proposed that molecules move like ideas, not the other way around. But Tarde was forgotten, and the physicalist understanding of the term gained the upper hand, followed by a further train of physical metaphors, such as "saturation" or "resistance". Latour (1986), inspired by the philosophy of Michel Serres, proposed its replacement with the word translation, calling attention to the richness of meanings associated with this term, only some of which are evoked in everyday speech. 1 Translation, he said, is a transportation combined with a transformation. Drawing from the same source, Callon (1980)  It is important to stress that the meaning of "translation" in this context surpasses the linguistic interpretation. It means "displacement, drift, invention, mediation, creation of a new link that did not exist before and modifies in part the two agents" (Latour 1993, p. 6), the two agents being those who translate and that which is translated. It is this richness of meaning, evoking associations with both movement and transformation, embracing both linguistic and material objects, that makes translation a key concept for understanding organizational change (Czarniawska & Sevón, 1996). It comprises what exists and what is created; the relationships between humans and ideas, ideas and objects, and humans and objects. 2 How can the notion of translation be useful in understanding globalization processes? Globalization is often depicted as a process that contributes to the compression of the world, a variation on the well-known definition by Robertson (1992, p. 8). But, as Sahlins (2001)  Robertson suggested that a more useful term could be glocalization or "telescoping global and local to make a blend" (Robertson, 1995, pp. 28-29).
Precisely because of its polysemic character, the concept of translation helps us to understand how this blend is achieved and how various types of connections are constructed around the globe. Translation thus understood means transformation and transference not only of utterances, but of anything. The concept thus redefined is meant to attract attention to the fact that any thing -an idea, a practice -moved from one place to another cannot reappear unchanged. To set something in a new place is to construct it anew.
Thus, translation immediately evokes symbolic associations; yet sociologists of translation emphasize that at the same time translation is inevitably material, because only a thing can be moved from one place to another and from one time to another. Ideas must materialize; symbols must be inscribed. A practice cannot travel; it must be simplified and abstracted into an idea, and thereby converted into words and images.
A similar translation is required in the opposite direction: no abstract model, no Best Practice description, no manual can guarantee that actions inspired by it will be identical.
The result of translation is always a change -a change in what was translated, and a change in the translator. The change may create improvement or deterioration, enrichment or impoverishment. Even copying machines do not produce copies identical to the original. Changes can be due to a faulty imitation, to a conscious adaptation to local circumstances, to the hidden hand of the past ("detritus" in Ossowska's terms -sediments in new institutional theory), but also because institutions are inscribed in machines (Joerges & Czarniawska, 1998), and machines -especially computers -play a central role in glocalization processes.
The travel of the idea of "operational risk" is one of many that could well illustrate such a process (Power, 2007), even if this text focuses on only a small part of this travel; in fact, it can be seen as an addendum to Power's story of its earlier peregrinations. It also shows, as every translator knows, that even when the translation remains limited to a linguistic practice, translation is a destabilizing operation. It destabilizes the text under translation, which is taken from its original cultural context and fit onto another, even if the context is in itself cross-cultural, as in the case of global accords.
Furthermore, the language into which the translation is made is destabilized, if ever so little, with every translation made -thus the need for the stabilizing role of dictionaries. One could claim that the text of the Basel II Accord was such a stabilizing dictionary (Power [2007], called it a "boundary object"). Yet a dictionary, no matter how perfect, does not guarantee the perfection of translation. It remains local, especially when words are being translated into practices.

Operational risk: A Swedish translation
When I began to question my Swedish colleagues about the connections between risk management and accounting, I was directed to a doctoral dissertation describing the history of public sector accounting in Sweden from the times of Gustaf Vasa to the present (Sandin, 1991). I consulted the work in question somewhat warily (in my early study of municipal reforms, I was told that they were initiated by the Vikings, which I believed for a while), and indeed, I found nothing on the topic. But at least two pieces of information captured my interest: the origin of Swedish accounting and the increased competence of accountants. Since the end of the 19th century, especially since the establishment of the Institute of Chartered Accountants in England and Wales, Swedish accounting and auditing were modeled after England's (Sandin, 1991, p. 43).
During the same period there was a wave of frauds, fal-sifications, and embezzlements. The solution was seen to be an increase in the competence of accountants and auditors (the Swedish Audit Society was founded in 1899) and a strong emphasis on impartial and competent auditing. Two central auditing organizations were created at the beginning of the next century, one for private companies and the other for public administration organizations. The notion of "risk" was not mentioned. The elimination of frauds and misappropriation was to be achieved by a stronger control: a retrospective rather than a prospective monitoring of economic activity.
A popular accounting textbook (four editions between 1983 and 1997) mentions "risk" on one page and defines it as a decision situation between complete certainty and complete uncertainty of its consequences. A situation is risky, then, when one knows the possible consequences of a given decision and their respective probabilities (G. Andersson, 1997, p. 38). Later in the text, two paragraphs were dedicated to "risk analysis" (pp. 195-196). Things have changed since 1983, however, in Sweden as elsewhere. A 1995 doctoral dissertation dedicated to risk and efficiency in interbank payment systems (M. Andersson, 1995) contained a simulation of system risk in payments.
Operational risk was seen as consisting of administrative risk and risk of fraud, and was commented upon in two half pages. But it was clear that risk management was becoming a central issue in financial institutions.
An explosion of the concept transformed this local, US finance (or military) invention into a global trait (Power, 2004).
As commonly claimed, the problematic liquidation of a German bank in 1974 required that the G10 nations form the Basel Committee on Banking Supervision, which reached its first accord in 1988: the so-called Basel I. Much attention was paid there to the notion of credit risk, soon to be complemented by market risk, and then, in Basel II (2004), by operational risk It has been argued that such events as 9/11 and rogue trading at Société Générale, Barings, Allied Irish Banks (AIB), and National Australia Bank convinced the committee that risk management extends beyond market and credit risk.
Indeed, when the translation of ideas into actions is well advanced, the actors involved feel a need to mythologize by dramatizing origins. It may well be that, in Operational Risk, Translation, and Globalization the reconstruction of the past, an event is chosen or invented because it is rhetorically convenient -a logical starting point for a story. Alternatively, the incidental and disruptive character of the initial events is stressed in order to demonstrate an incredible touch of luck in the timely arrival of the idea. Both types of memories serve the same purpose: to tie, meaningfully, the arrival of an idea to present problems experienced by people in organizations or attributed to the organizations. There is often an attempt to portray the process as functional; this idea was spotted and adopted because it served well in resolving a specific difficulty or in creating a new opportunity in situations of stagnation. Although the label, "operational risk", has existed since the beginning of the 1990s, Power (2007) has noted that the banking community tends to relate it to the collapse of Barings Bank in 1995.
The Basel Committee defined operational risk in 2001, so that Gunnar Wahlström was able to interview Swedish bank managers that year (Wahlström 2006(Wahlström , 2009a and again in 2005 (Wahlström, 2009b). As the title of one of his articles suggests "Risk in practice"( 2009a), he was interested in risk management in banking. There are no Swedish textbooks dedicated to risk management and accounting, and textbooks dedicated to risk management tend to treat it in general terms (e.g. Hamilton, 1985Hamilton, /1996. Perhaps because Swedish students -and bankers -read English, there is no need for linguistic translations of texts dedicated to such issues. There is only one reference to a Swedish text in Wahlström's three articles, although he quotes many Swedish and Nordic authors. In both his stud- was also seen as a valuable complement to credit and market risk measurements, touching such potentially threatening areas as fraud and embezzlement, which often create crisis, but were not previously covered by risk measurements. 4 Additionally, the procedure was seen as enhancing the banks' relationships with other industries by improving clients' understanding of the bank's situation and by providing clients with a model of a successful procedure that includes operational risk -a procedure that they may consider imitating in their own risk management. The critique of the notion of operational risk began with the opinion that Basel II's definition was vague and abstract, thus making its proper measurement deeply problematic. The second criticism was more psychological, placing in doubt employees' willingness to expose their failings. 5 All in all, Wahlström (2006) concluded, the interviewed managers were strongly in favor of the innovation, and he attempted to explain this somewhat surprising (in the face of criticism) enthusiasm: in the society and is manifested in a conviction that it is possible and appropriate to measure risk (Wahlström, 2006, p. 512).
Wahlström is symmetric in his approach to the field studies, in that he does not issue a priori judgments.
But in communicating the results, he is clearly referring to the critical and sociological school of accounting studies.
In addition to collecting the interviewees' opinions about Basel II, Wahlström also asked them what they perceived as being the greatest risks in their work (Wahlström, 2009a). By his own admission, the an- (…) Instead, the interviewees talked much more about risks that they could not measure" (Wahlström, 2009a, p. 291). Indeed, senior bank managers agreed that the greatest risk areas are those that defy quantified measurement, and pointed out that their most challenging task is to solve unanticipated problems. This second statement corresponds well to other studies of risk, in that it highlights the importance of "well-practiced improvisation" (Czarniawska, 2009 After the Basel II Accord was accepted, Wahlström returned to the four banks and conducted a new series of interviews. As before, opinions were divided into positive and negative, and as before, there was general support for Basel II. The positive opinions stressed the concordance of suggestions in the accord and actual banking practices (which may partly explain the tendency among Swedish organizations to implement the EU's and other international regulations well in advance; see Jacobsson, [1993]). The interviewees were also pleased by the fact that the accord permitted the banks to use their own measurement models, whereas the requirement for a measurement (not an assessment) of risks imposed a desirable uniformity on the banks. The criticisms revealed large gaps among the groups, however: managers with operational functions (usually older) and staff specialists in risk management (usually younger, with higher levels of education). The staff people, not surprisingly, thought all was well with operational risk measurement, whereas the operational people took a relatively critical view of the younger group, claiming a fissure in "our risk organization, with its PhDs and statisticians on the one side, and the managers who run the bank on the other". As one operational manager said, risk management "forms its own tradition in the theoretical world and at the universities. And soon there will be just three people in each bank who really understand the rules and can explain them" (Wahlström, 2009b, p. 61). Managers involved in running the operations noted that implementation of Basel II was costly (indeed, there are many IT companies that started to specialize in appropriate systems and software), but, more importantly, given the abstractness of Basel II, its relationship to reality is doubtful. Additionally, they saw Basel II as supporting a tendency toward centralization, which the more decentralized banks saw as a threat. As always, when ideas are translated into practices, they involve people and objects, identities, and computers.
The critics were well aware of the need for local translations. They pointed out that the accord itself was vague, but was considerably clarified when the Swedish Financial Supervisory Authority introduced its interpretation, based on a dialogue with the banks. As one CFO said, however, "We are afraid that the Swedish Supervisory Authority will interpret the regulation more strictly than the supervisory authorities in other countries. So we are afraid that there may be competitive disadvantages for Swedish banks with non-Swedish banks, including those with branches here in Sweden" (Wahlström, 2009b, p. 63).
So now I inspect another local translation -this time in Poland. The starting definition of risk is the same as in G. Andersson (1983Andersson ( /1997 that revealed the percentage of the four components of operational risk: 64% procedures, 25% people, 2% systems, and 7% external events (Gospodarowicz, 2009, p. 270). The chapter proceeds by listing various typologies of operational risk (including the Basel II typology; see Endnote 2), and states that its measurement is extremely difficult. Another statement, which was meant seriously, I assume, maintains that the measurement of operational risk is 80% art and 20% science, given that "the losses that determine the level of operational risk are under a strong influence of human behaviors, which are difficult to foresee and often unrepeatable" (Gospodarowicz, 2009, p. 273). Measurement methods can be qualitative (scenario analy-sis, Delphi method) or quantitative, the later dividable into top-down and bottom-up (English in original).

Operational risk: A Polish translation
Another division differentiates between internal and external methods, and yet another between basic and advanced, the latter to be used only after permission of the supervisory authority has been granted. The chapter continues by presenting four methods of growing complexity (the most advanced is Loss Distribution Approach), and concludes that banks will certainly take several criteria into account when determining the method to be used. The last paragraph mentions a growing supply of IT systems specializing in the measurement of operational risk, as exemplified by Canadian Algorithmics (their product is called Algo-Suite) and Viennese BOC Information Technologies Polish banks in the years ahead (Lewandowski, 2004).
"Sound practices" have been rendered in Polish as "Best Practices".
After the 2007 tax reform and the Banking Supervision Committee (since then incorporated into Financial Supervision Committee) presented its interpretation of Basel II (I was unable to establish if it happened in a dialogue with the banks, as it had in Sweden), Bancarewicz (2007) wrote an article that she began by quoting an assessment of risk (25-30% operational risk, 65-70% credit risk, and 10% market risk). The original article by Lenczewski, Martin and Niedziółka (2005) refers only to "research results", however, without specifying who conducted this research or where. Bancarewicz (2007) quoted the same percentages of causes as Gospodarowicz (2009) did in his chapter, but increased their visibility by presenting them graphically. The main part of the article, to quote its English abstract, "shows relevant difficulties and challenges that a bank may come across while collecting loss data and modeling operational risk" (Bancarewicz, 2007, p. 96). The analysis is purely speculative; in two places unspecified "other European banks" are mentioned.
Lenczewski Martin and Niedziółka's (2005) article presents some research results, but primarily speculations and assessment, as their study was done immediately after Basel II and before the interpretation by Banking Supervision Committee. They predicted some of the same problems as Wahlström did, albeit some in the opposite direction. Different local translations may cause problems for international banks in Poland, not the other way around (although it is not spelled out that Polish interpretations are likely more tolerant). The classification of events included in operational risk may be a problem: the same events can be classified as market and operational risk, they can be wrongly monitored, and their consequences may be difficult to ascribe to an appropriate category. And, as in Sweden, many speculations concern the impact of operational risk on the required level of regulatory capital; but, unlike in Sweden, the expectation was that the measuring of operational risk would raise rather than lower this level. I have finally located in "Risk in accounting"an ambitious co-authored work edited by Anna Karmańska (2008) -a small study that can be compared to Wahlström's. The entire book has nine authors, runs for 535 pages, and can probably be characterized as a handbook. It has no assignments and cases like other textbooks typically do. The 17 chapters are primarily encyclopedic, although each begins with a motto borrowed from a philosopher, a novelist, or even Yogi Berra, and some contain hypothetical examples. There are 234 pages of lists, organized with the help of Arabic numbers, Roman numbers, capitals, low caps, bullets, and dashes; 141 pages contain tables, many with lists. Only 24 pages contain equations, and as many have graphs or figures.
The survey of interest to me was presented on the last four pages of the text, with questions addressed to four top-level managers, who were well acquainted with both financial accounting and management accounting procedures. They were shown a risk report form, constructed according to suggestions from the authors of the volume, in which risk was separated into several categories, relating primarily to market and credit risk. The first question to these four managers was whether or not such a report was needed for financial analysis. The general answer was "yes", with the specification that if the report were to be correct, it had to remain internal information, and it would be of more use to large and middle-sized companies than to small companies. The second question concerned the suggested structure of the report, which achieved disparate responses. Two of the managers liked it, one was uncertain, and the fourth thought that open questions invited politically correct answers, although the same person admitted that closed questions could be difficult to formulate. The question concerning possible arenas of use raised some anxiety about the report becoming obligatory; but three persons saw it as a useful source of managerial information. Asked if some parts of the report should be better developed, the respondents protested, which can probably be explained by their answer to the next question, in which they estimated the preparation as time-and effort-consuming, at least the first time around. In their opinion, such a report should be prepared by their finance department. When asked if they had competent personnel who could accomplish the task, the only manager from Operational Risk, Translation, and Globalization a large company answered positively, yet added immediately that someone that competent should be occupied with more creative tasks. The other three managers answered negatively, as they considered their companies too small to include such personnel, the lack of which would require a preparation time of between two months and two years. In addition, three of the mangers were convinced that if the report had an external function, it would be no doubt manipulated.
The editor ended the volume by emphasizing the complexity of risk reporting and suggesting that a great many empirical studies were needed in the future. The fact that the sample used for the survey has been so limited is easier to understand in the light of the fact that Cap Gemini failed to conduct "Basel II Survey" among Polish banks in 2004. The response rate was so low that it was impossible to draw any conclusions from their study. The reporter who wrote the article reporting this failure asked several top managers for reasons (Gamdzyk, 2004). Those from other industries suggested that banks did not have and still do not collect appropriate data, and that bank managers were afraid of spending money on uncertain investments.   (Jacobsson, 1993, p. 113;my translation, BC).
Indeed, studies from various times and of various areas show a similar picture: the Swedish public administration is always au courant with the newest fashions in what was previously called administration, and is now called management. As a "negotiated economy" (Hernes, 1978), it was and is in constant dialogue with the private sector. For many years it was a global "fashion leader", proudly presenting "the Swedish model" of a welfare state to visitors, it admitted its demise   Because of the dramatic change of the political regime and the consequent changes in running the economy, there are no "old, experienced bankers". Furthermore, one may suspect a common positive perception of abstract models and numeric calculations across generations, due to an anti-pragmatic attitude, which I called "merciless idealism" in my city study (Czarniawska, 2002, p. 119). One of its elements is "trust in numbers" (Porter [1995]; but also "trust in models"; Power [2007, p. 120]), albeit with a local twist. Numbers were commonly manipulated during the socialist regime, but faith in "correct numbers" remains -the correct numbers guaranteed by a non-ideological science. Rottenburg (1994)  But the contacts and mediations do happen, and researchers are often asked to join or lead official investigations.
Between 1972 and1980, I served as a methodological consultant and researcher in a research program "Managing enterprises -participants in the consumer goods market" (Beksiak, 1978), in which an extensive field study of actual management practices was conducted, probably for the first time. Field studies in Poland have not vanished, for Polish and foreign researchers are conducting organizational ethnographies (Kostera, 2011). But such studies, being unquantifiable, are seen as being of no use in an official context. "Calculative pragmatists" (Power, 2007, p. 121) seems to be an empty category, or at least a not-yet-located group. The label is a good fit with the older operation managers interviewed by Wahlström. 7 Sweden has a long tradition of field studies and studies of practices, and in this aspect does not have to imitate UK. A critical and sociological take on accounting, however, is clearly a UK influence. Yet I am not sure if Polish accounting scholars are at all aware that accounting can be a social science 8 (neither Accounting, Organizations and Society or Critical Perspectives on Accounting are to be found in Warsaw University Library). As I see it, "socialist monism" has been replaced in the official discourse by "capitalist monism"; the top-to-bottom" approach in centrally steered initiatives remains, as does the role of the researchers as translators of top leaders' intentions to the wider public.
However, this analysis must not be read as an eu-  Robson, to whom translation is "the process through which often pre-existing accounting techniques, and their associated roles, are articulated discoursively" (Robson, 1991, p. 550 (Power, 2007: 121