Accueil Revue Numéro Article


2015/3 (Vol. 18)

  • Pages : 90
  • DOI : 10.3917/mana.183.0261
  • Éditeur : AIMS


Votre alerte a bien été prise en compte.

Vous recevrez un email à chaque nouvelle parution d'un numéro de cette revue.


Article précédent Pages 261 - 265 Article suivant

The study of institutional change over a period of history involves several theoretical and methodological issues. In proposing a commentary on our article published in Vol. 17 n°3 of M@n@gement (Peton and Pezé, 2014), Daudigeos, Boutinot and Jaumier have provided an opportunity to open up the debate, and we thank them for that. Their commentary concerns two major points: a methodological question concerning the choice of data in a historical research context, and a resulting reflection on the consideration given to collective representations in analysis of institutional work.


Before replying in concrete terms to the points raised by Daudigeos et al., we wish to return to the objective of our article and clear up one matter they raise. Our study starts with reflection on the different institutional pillars. While Scott (2008) stresses the importance of all three pillars (normative, cognitive and regulative), the regulative pillar is rarely studied in the literature on change. It finds itself relegated to the role of a ratifier of changes driven by the cognitive and normative pillars, and has generally been considered static and treated as the same thing as the rule produced by the State (Scott, 2008; McCann, 1994). The case of Faute Inexcusable, an institutionalised practice introduced in 1898 and still in force today, reveals the complexity of a regulative pillar that has a real, and very specific, institutional dynamic.


We can only agree with Daudigeos et al. when they stress the importance of paying attention to interaction between institutional pillars in institutional change. Our article focused mainly on the regulative pillar, such that we do not propose any contribution regarding interactions between pillars and their mutual effects – that was not our research project. However, our contribution to the understanding – and a form of intellectual rehabilitation – of the regulative pillar paves the way for a broader research agenda explicitly aimed at exploring the interaction dynamic between the pillars, without claiming to answer it comprehensively [1][1] This brief overview of our approach is given to avoid....



We agree with several of the remarks made by Daudigeos et al. First, the fact of not restricting the organisational field and the actors involved ex ante; second, the usefulness of taking a publication that is central to the field to capture change over time in the field’s boundaries, and thus remain sensitive to its “plasticity” [2][2] This is the methodology used in many longitudinal studies.... Understanding change in the field through the study of a central publication is indeed both very informative and relevant.


And yet Daudigeos et al. state that our methodological choice, i.e. our concentration on legal and judicial data, could result in certain actors being ignored. They argue that this choice – which is one of the contributions our research makes to management science, highlighting the informative value of legal and judicial data that are still rarely used in our field of research – enables us to show the work of hitherto ignored actors, but also restricts the range of actors observed. To illustrate their point, Daudigeos et al. argue that the choice of a central publication from the field would have brought other actors to the fore. They give the example of the findings of an empirical study drawing on the archives of the trade journal Le Moniteur des Travaux Publics et du Bâtiment since 1902, and note that insurance companies played a role in the original law of 1898 (which gave rise to the concept of Faute Inexcusable) and its future changes. This in their opinion completes our “rather limited vision of the components of civil society” which they consider as a bias generated by the nature of the legal and judicial sources used. While the additional information provided by Daudigeos et al. may indeed identify an additional actor which played a lobbying role in the origins of Faute Inexcusable and its evolution (at least between 1898 and 1941) [3][3] However, the relevance of using a corpus that starts..., there are several responses to this critical reading. First, it should be remembered that the original aim of our research was to offer an initial interpretation of the regulative pillar’s internal dynamic, not to study an institution in all its complexity. Second, although we emphasise the usefulness of studying original data from legal and judicial sources, we also based our study on more contextual research [4][4] We drew on other sources of data, such as the work.... And finally, the illustration proposed by Daudigeos et al. raises the question of studying a field in all its completeness without finding a totally satisfactory answer. There is apparently no simple answer to the question of the right methodological choices for a complete, in-depth study of a field encompassing all its length and its multiple dimensions. This is why, as well as aiming for completeness, selection of the data source must primarily respond to the issues of the research subject. It partly explains why so little research has so far attempted to study all three institutional pillars at once. This once again raises the question of the central importance of the data source used for research. The example of Le Moniteur reveals the difficulty of conceptualising and operationalising this idea. Le Moniteur may be the trade journal with the widest circulation in France, but we would be more cautious than Daudigeos et al. over their assertion that it has the status of “a publication that occupies a central position in the field and whose development of the content thus reflects the evolution of the organizational-field boundaries and thus the changes in the organizations and interest groups that populate it” (p. 255). For one thing, Le Moniteur is intended for members of the construction and civil engineering industries – which is similar to the limitations Daudigeos et al. consider to exist in our choice of data sources, and the ex ante delimitation of the actors studied. For another, they state that it provides information on a very wide range of subjects (all related to the construction industry). Given the diversity of themes covered, then, how can this publication occupy a “central position in the field” of Faute Inexcusable specifically [5][5] Another very stimulating discussion could be started...?


More generally, this point relates to recurring questions in analysis of an institutional field: who are the actors whose institutional work is considered analysable? The actors able to express their opinions in the chosen publication? But in that case, what about other business sectors or “silent voices” that do not have access to that publication? Research into agency, particularly its different forms, could develop genuinely useful reflection on these questions: is it necessary to have projective agency to carry real institutional change? Previous research (e.g. Leblebici, 1991) has already underlined the role of peripheral actors in institutional change: so how can we measure their institutional work through one “central” publication?



The second criticism made by Daudigeos et al. is that our focus on the regulative pillar leads us to neglect the continuous influence of the cognitive pillar and thus propose “a biased interpretation of causal links” (p. 257) concerning Faute Inexcusable. In fact, we do not study the change in an institutional practice for its own sake, but in order to capture the complexity of the regulative pillar and its dynamic – hence our decision to focus more on that pillar and political institutional work. This comment – which is also valid for many studies on institutional change – has the merit of questioning the relevance of studying the regulative pillar without an associated study of the cognitive pillar. Yet Daudigeos et al. acknowledge that we do identify the interactions between the cognitive and regulative pillars during one of the periods identified: in the end, their main criticism is that we did not do so systematically for all the periods. More specifically, Daudigeos et al. question our empirical analysis, proposing a “counterpoint” (p. 257) to our explanation of the origins of the turnaround in case law from 2002 onwards by arguing that “the contestation of the compensatory amounts put forth by Peton and Pezé is only the manifestation of a much deeper contestation of the principle of the 1898 law” (p. 257) [6][6] On this point, Daudigeos et al. analyse the change.... But this is more of a confirmation than a counterpoint to our analysis: it was definitely a misalignment between the regulative and cognitive pillars that brought the Cour de Cassation judges to change direction in its ruling of 2002. We agree that the cognitive pillar “influences all steps of institutional change and cannot be ascribed to the sole role of a trigger” (p. 257). Once again, to avoid any misunderstanding, it was never our aim to confine one pillar to a precise role in the absolute (for instance, the role of trigger) – instead, we have written an empirical study of a period of change in the concept of Faute Inexcusable, confirmed by Daudigeos et al., which identifies a type of contingent relationship at a given point in time in the case studied.


To extend the logic applied by Daudigeos et al., the normative pillar (which, surprisingly, they do not mention) should also be central to the analysis, among other things leading us to question the role of materiality or artefacts in understanding institutional change (and thus potentially opening up other theoretical discussions)! This discussion thus underlines another level of complexity in the analysis of historical change in institutions: detailed examination of the interactions between the three pillars, which doubtless do not play the same role with the same intensity throughout the institution’s lifetime. Currently, there are few studies that analyse the three pillars of the institution simultaneously (two notable exceptions are Caronna, 2004 and Maguire and Hardy, 2009). The many interesting studies on the concept of institutional logics offer a more integrative understanding of institutional factors. Logics can be defined as “the socially constructed, historical patterns of material practices, assumptions, values, beliefs, and rules by which individuals produce and reproduce their material subsistence, organize time and space, and provide meaning to their social reality” (Thornton and Ocasio, 1999, p. 804). This approach would certainly be more in line with the expectations of Daudigeos et al. regarding understanding of change in a field or institution. Change, in this view, results from struggles, oppositions, integrations or hybridations of the different logics that structure the institution: complexity thus brings about change (Greenwood et al., 2011). However, this approach does not appear to provide any answers to the question of the mutual interactions and influences between the various components of the institution, which can be examined in a more differentiated way through a pillar-based approach. One of the contributions of this type of analysis is that it focuses attention on one of its dimensions – like our study, which shows the specific dynamic of the regulative dimension and can thus propose an initial understanding of the misalignments that may appear in an institution. Our analysis is thus situated upstream of the conclusions of Caronna (2004), who notes that institutional change is due to institutional alignments and misalignments. By better understanding the dynamics internal to the pillars, we offer a better understanding of the misalignments that are potential sources of institutional change. The regulative dynamic highlighted in our study, for instance, enhances understanding of the institutional work of defence or destruction going on inside institutional change.



In the conclusions to their commentary, Daudigeos et al. sum up their critique as follows: “the absence of a systematic consideration of the changing boundaries of the field and collective representations leads to the actors and their rationalities being arbitrarily defined, and the role of power struggles in institutional change being overemphasized” (p. 257) This rejoinder resituates our research in its intellectual context and the rich and varied corpus of literature on institutional change. We fully agree with many of the comments and remarks made, and extend their analysis by showing that the choice of a central publication does not in itself provide a satisfactory answer to the questions of which actors to study and which to leave in the background. Their objections are not so much a limitation to our study as a plea for further research to tackle the complex methodological challenges of combining the different levels of complexity mentioned.


Historical or processual research over long time periods involves methodological issues that are accentuated here, since studying an organisational field of institutionalised practice is a complex undertaking. The choice of the data source, as is often the case in processual research, restricts the analysis just as much as it brings out the subtleties of interaction between some of the actors involved. This, we would argue, is one of the reasons for the low number of empirical studies that analyse all three dimensions of the institution simultaneously and in depth [7][7] The question of causality in sequential studies could....


We believe that our study of Faute Inexcusable offers an initial exploration of the regulative dynamic which, as we suggested in the avenues for research at the end of our article (p. 173-174), needs to be questioned and extended to encompass other actors, but also other sources of law (such as legally binding collective agreements). Our study of Faute Inexcusable provides conceptual elements for a detailed analysis of the regulative pillar and its interactions with the other pillars, the scale and nature of those interactions being neither identifiable ex ante nor necessarily stable over time. This interpretation of institutional change – which may also hold good for dynamics of institutional maintenance – is very promising, and we can only call for more long-term empirical studies drawing on several sources of data.



This brief overview of our approach is given to avoid any misunderstanding: Daudigeos et al., at the end of their first paragraph on the presentation of our article, write that “[Peton and Pezé] ‘highlight a regulative pillar that is ‘ahead’ of the cognitive and normative pillar’ within institutional change (page 174).” This extract from our article is not in fact a stated contribution, but a hypothesis we suggest as part of the directions for future research.


This is the methodology used in many longitudinal studies in neo-institutionalism which are founded on systematic study of the press or some specific publication.


However, the relevance of using a corpus that starts in 1902 to argue that “insurance companies became involved in the political battle around the 1898 law” (p. 255) is debatable, particularly as the corresponding process began in 1880 (Bienvenu, 1938). Other sources, such as the archives of the said insurance companies, could cast more detailed light on their exact role.


We drew on other sources of data, such as the work of historians specialising in occupational health and safety (Boisselier, 2008; Buzzi, et al., 2006; Machu, 2009; Rosental, 2008; Seillan, 2009), who do not specifically note the role of insurance companies in the birth or subsequent evolution of Faute Inexcusable.


Another very stimulating discussion could be started here, on the question of the unit of analysis: which institution is actually being studied? Can Le Moniteur be considered central for analysis of the institutionalised practice of Faute Inexcusable? Does the study by Daudigeos et al. in fact examine the field of occupational accidents and liability for those accidents? If so, it would arguably have been better to choose a publication such as the journal Travail & Sécurité, published by the French national research institute for prevention of occupational accident and illness (INRS), which has the added advantage of not being restricted to the construction and civil engineering industries, but the disadvantage of only being published since January 1949.


On this point, Daudigeos et al. analyse the change in the “conception of workplace accidents” and find that it oscillates between “universal protection of workers and massive investment into technical prevention” in 1941, and being “about individual responsibility” in 2002 (p. 256). This gives the impression they are studying the more general institutional field of occupational (workplace) accidents and how reparation is made, rather than the more precise and clearly delimited (albeit embedded in the broader field) field of Faute Inexcusable. This vagueness regarding the exact definition of the institutional field studied is not unproblematic, because there is a risk that Daudigeos et al. will use the cognitive pillar from a different field to engage in dialogue with the field concerned by our study.


The question of causality in sequential studies could be added to this very interesting debate. Apart from interaction, characterisation of mutual influence remains problematic today. Detailed understanding of the causes of an institutional change, especially the succession of phases of maintenance and contestation, is a methodological challenge. The approach via institutional struggles has brought out the existence of the unstable balance beneath an apparent institutional status quo, but it would be interesting to have in-depth studies of the shift between a stable and an unstable balance.

Plan de l'article


Article précédent Pages 261 - 265 Article suivant
© 2010-2017