Interview with Emmanuelle Marchal
Anna Schneider & Qamar Ali (University of Innsbruck, Austria)
Emmanuelle Marchal is a CNRS Research Professor at the Centre de Sociologie des Organisations, (Center for the Sociology of Organizations) in Paris. Her research focuses mainly on the sociology of labor market and the sociology of valuation. She has carried out many qualitative and quantitative surveys on the functioning of labor market, emphasizing the role of devices, intermediaries, trust, recruitment channels, hiring and screening processes.
Schneider/Ali: Please tell us how EC has been beneficial to your research work so far?
Marchal: To answer your question, it is important to know that I was in contact with the founders of Economics of convention School (in short EC) even before the publication of the founding work of 1989. I was a researcher at the CEE (Centre d’études de l’emploi) when François Eymard-Duvernay took over as director of the Centre and brought Laurent Thévenot in as scientific director. (CEE has recently been integrated into the CNAM.) As a result, I have been “imbued” with the theses of EC, as they have developed. I was able to benefit not only from the reading of the work done on job classifications, companies, public policy intermediaries, and The economies of worth, but also from the daily discussions to which they gave rise. Beyond this first circle of conventionalists, I was also able to benefit from the contacts they were developing with the GSPM (Groupe de sociologie politique et morale) founded by Luc Boltanski, and with the CSI (Centre de sociologie de l’innovation) where Bruno Latour and Michel Callon were developing their own theses. It was all very exciting.
I started working with François Eymard-Duvernay at the beginning of the 1990s, by carrying out a survey on the introduction of New Public Management, in a social housing management organization (in French Office HLM). We were particularly interested in the impact of how changing management styles affects the perceived value of people. Many public servants, previously highly rated, were devalued overnight. Hiring methods were completely revised, abandoning competitions and knowledge tests in favor of private agencies and graphology. This was accompanied by a complete revamp of the organization chart and job titles, as well as an introduction of new methods to evaluate staff and activities. What was evident from our observations, was that the outcome of the individual’s performance evaluation could be very different depending on the conventions valued and chosen by the organization’s managers to evaluate them. This is what we called “skill conventions” (in French, conventions de competence) in the book we published together in 1996, following inquiries done in recruitment agencies. This book marks the beginning of a long collaboration with François Eymard-Duvernay and with the economics PhD students he supervised later in Nanterre University. We developed a typology of “Ways of Hiring” (Façons de recruter is the title of the book), to highlight that each of them carries its own specific interpretation of competency. In some situations, skills are planned: they are seen as pre-established and measurable, independently of the context of its use. On the contrary, they can be negotiated during the hiring process between actors in such a way they are treated as the result of a process. The second continuum goes from cases where skills are considered as embedded in collectives (organizations, networks, workplaces) to cases where they are individualized, as if they were carried by the applicant alone. Depending on the ways of recruiting and screening methods used, the stage of the process and the type of actors involved (when there are several), different conventions are mobilized. As a result, different applicants are valued, selected or excluded. That is why we consider hiring activities as valorization / devalorization processes.
I developed this point in my last book (Les embarras des recruteurs), where I summarize the work I have been doing for 25 years. Starting from the classic distinction between remote coordination, which is highly equipped, and proximity coordination, which more often takes place face-to-face, I highlight the dysfunctions of the labor market in France. I show how recruitment and selection methods can contribute to penalizing some people, and how they maintain long-term unemployment and exclusion.
Schneider/Ali: Give us a brief overview of your research journey in recruitment, work in organizations, and labor markets so far.
Marchal: As a sociologist, I value empirical investigations. So I have tried to enter into numerous collaborations and to use all forms of methods of inquiry (observations, interviews, textual and statistical data analysis). My very first investigation was based on in-depth observations of professional consultants’ activities in private agencies. I was fortunate to be able to follow them in their daily work, including attending interviews with job candidates. This position gave me exposure to the uncertainties they were facing. At the same time, I worked with a colleague from INSEE (The National Institute of Statistics and Economic Studies) on the French LFS (labor force survey), to show how recruitment channels were distributed, according to the population who had found employment. I was able to work there again later with Géraldine Rieucau, in comparing results from French and Spain. It must be said that this was the only statistical source available on the topic in the mid-90’s. Thereafter, we became interested in job advertisements and undertook historical and comparative analyses (in France, Great Britain, Spain) to give an account of how the formulated requirements and the language used varied according to these labor markets. Following that, we initiated a major survey, for which I designed the questionnaire. This OFER survey was conducted by the Ministry of Employment in 2005 and 2016 among 8 600 establishments. It makes it possible to trace very precisely the last recruitment method, and shows the great variety of research methods and evaluation of job offers and skills. I worked with Christian Bessy on recruitment channels and with Guillemette de Larquier on the impact of hiring methods on the selection of candidate profiles. I am currently exploring the results by focusing on how « non-qualified » staff (low skilled workers) are recruited. How do you select candidates whose qualifications are not recognized? What is valued in this particular market?
Previously, I have also participated in the evaluation of public recruitment policies, by conducting surveys at Pôle emploi (The French public employment service), among private and public market intermediaries and large companies. As a result, I contributed to the evaluation of the relevance of the potential introduction of the compulsory anonymous Resume and of a particular recruitment method which requires the candidate to simulate the tasks involved in the role (MRS in French). I was also interested in and conducted research on the historical evolution of methods and the introduction of personality analyses in recruitment, via graphology and tests, as well as their criticism. Three years ago, I started a new project on recruitment fairs with Delphine Remillon and Géraldine Rieucau: At a time when most companies are focusing on the all-digital world, what is at stake in these brief face-to-face interviews?
Schneider/Ali: Please tell us what particular aspects of EC you are currently referring to in your ongoing research projects.
Marchal: At the moment, I am conducting an exploration of grievances related to cases of hiring discrimination, and filed through the Rights’ Defender. The question of discrimination has already been addressed from a conventionalist perspective, by Ariane Ghirardello, through her PhD. In the cases treated by the Rights’ Defender’s lawyers, I explore the elements that claimants use to support the presumption of discrimination, and that employers use to defend themselves. In an ideal situation, it should be possible to decide: yes or no, did the employer use a discriminatory criteria to reject a candidate? If not, he must be able to provide the justification of his choice. But there is a lot of uncertainties around the definition of discriminatory facts. What is at issue is the identification of the pivotal factor of the recruitment decision which often involves many actors and many tools. Sometimes, candidates are not unknowns, because they have previously been employed by the same employer; Disagreements over the “worth” of workers (applicants) are frequently at stake; The assessments differ not only between the claimants and respondents in the same situation, but also from one step of the hiring process to another. Judgments may be rightly based on the candidate’s personality or motivation, placing the lawyer in a delicate position to determine whether his rejection appears legitimate or not. The complaints can often go far beyond the issue of discrimination, in addressing the fairness within the hiring process.
Schneider/Ali: What future issues should/could EC scholars address? Which do you think are the most pressing questions related to labor markets EC scholars should address?
Marchal: It seems important to me to follow the evolution of the labor market by highlighting the different ways in which job providers and job-seekers manage the coordination and valuation of themselves on the markets. A good way to do this, is to focus on devices that enable the evaluation and matching of market actors in the market. Most of the time, these tools are designed by employers and market intermediaries (and not by job seekers). What does this « valorization power » (Eymard-Duvernay) produce? This question arises both when entering organizations and when moving internally within them. What are the conventions embodied in the evaluation tools used in any one particular company?
The use of the Internet to recruit may permanently lead to favoring very general and transversal matching tools, with the aim of circulating the workforce over large areas. This is the trend of the public employment service in France, which favors standard and universal markers for recording and classifying job offers and applications. This choice leads to devaluing the markers embedded in local markets and professional markets that would allow many candidates and employers to meet up: locally known schools or companies are not recognized by the standardized tools that devalue them. This fuels recruitment difficulties and long-term unemployment. On the other hand, the role of intermediaries specialized in sectors of activity, trades and localities could be studied further to highlight their contribution. The same is true of networks that not only circulate information, as Granovetter pointed out, but also trust. As economic sociology shows us, no market works without trust devices, “personal devices” or” impersonal devices” (Karpik) or without “investment in forms” (Thévenot).
Another example is the algorithms that some start-ups are proposing to use to recruit. What assumptions and what knowledge are incorporated into these devices? I think the idea is always to predict competence, just like what the psychotechnicians dreamed of doing at the beginning of the 20th century. If an individual succeeds in a job, there are hundreds of data about him or her (tastes, hobbies, personality, network and background) that can be recorded and processed. Would an individual with a similar profile succeed in the same job? The tool is innovative, but the approach is not. Cognitive sciences, ergonomists and the psychodynamics of work have shown to what extent skills are situated, collective, distributed and evolutionary. There are many ways to accomplish a task and it is important to remember that competency judgments are not intended to reveal or measure skills, but to agree on what makes the workers and activities valuable. A lot can be learnt by looking into how other scientific disciplines approach our own research topics. That said, we should take a closer look at the assumptions embedded in the invention of algorithms. The rise of behavioral sciences in public tools and polices is worrying, from this point of view.
Schneider/Ali: In relation to this, and with regards to your multiple use of both qualitative and quantitative methods, how would you describe your methodological stance towards investigating social phenomena from an EC perspective?
Marchal: I don’t think that there is any specific « EC method », but you are right to underline that adopting an EC perspective has methodological incidences. How to account for the actors’ reflexivity, concerns and constraints? How to take into account the tensions generated by the possibility of referring to several orders of worth and justice? Personally, I subscribe to the principles of the pragmatic sociology developed by Luc Boltanski. I have also been very influenced by the work of interactionists, ethnomethodologists and by the reading of Alfred Schütz. Finding the right distance to the field and to the actors, implies to respect their critical capacities without trying to impose our own point of view. The difficulty is also to account jointly for what stabilizes the social world and what transforms it, as Nicolas Dodier explains well. Observing the activities of actors who are at the crossroads of several worlds is often a good entry point into the field. This is the case, for example, of market intermediaries, of actors who are at the borders of organizations and develop relationships with clients or users, of spokespersons for social movements or State inspectors. This is the case for all those who find themselves at the crossroads of often contradictory requirements, whose activities are easily exposed to criticism. This is also the case of actors who have more distance to what they are doing, because they have to implement an innovation or to face an organizational change, because they have just been hired or have been displaced, because their routine activities are interrupted by something. All of them have interesting things to tell us about what is happening in the social world.
Using quantitative methods is more complicated, because we cannot ignore the work done by Alain Desrosières and by Laurent Thévenot on the history of statistics and nomenclatures, on the choices made when codifying and classifying data. We know that any statistical data incorporates conventions. They incorporate ways of presenting reality that have led to the exclusion of other versions of the world as explained by Espeland and Stevens. It is tempting to leave it there and to focus only on the analysis of their construction: how do individuals go about quantifying what is qualitative? But we also know that quantitative data has more credibility than any qualitative data, especially with decision-makers.
Like all actors, we must make compromises. This is what we tried to do by using the text data processing software created by Francis Chateauraynaud (Prospero). The objective was to quantify the content of the advertisements, without sacrificing the importance of the wording chosen by employers to write their job offers. The vocabulary of job titles, for example, changes over time and is formulated on several registers that should not be eliminated during operations of quantification. Some statistical methods are specially designed to smooth the transition between qualitative and quantitative data, by mastering the intermediate steps. It is the case of the multiple correspondance analysis (MCA) used with Guillemette de Larquier to explore the OFER Survey. We propose an inductive typology of screening process used in hiring. The method enables us to measure the distribution of the four screening processes identified and to measure their impact on equal opportunities.
Schneider/Ali: Regarding your interest in the recruiting of “non-qualified” staff: how can an EC perspective explain the “revaluation” of immigrants’ skills. In other words, how can the EC explain why physicians or engineers end up working as taxi drivers in their host country? Which “skill conventions” are at work here?
Marchal: It is difficult for me to answer this question on immigrant workers because I have not done any specific work on this population. I tend to think that the problems are different in between highly qualified people, like the engineers or physicians you mentioned, and unskilled labor. In the first case, it is the (non)recognition of diplomas by the host institutions which is at stake. For the latter Annalisa Lendaro and Christian Imdorf use the EC framework to show how intermediaries maintain the segregation of immigrant women by assigning them to the domestic labor industry. Immigrant women and mothers appear as cheap labor, having built-in female and traditional skills, which are not recognized as having a market value.
As for me, from what I can see in the OFER survey, is that more than half of recruiters do not recognize the value of unskilled workers diploma. And they do this deliberately whether they are immigrants or not. Being a woman is an asset to be hired as a home helper but represents a handicap to work in the construction sector. You must be young to work in the hotel trade, restaurant industry or in commerce, whereas very few older workers would be hired. The latter are confined to cleaning jobs where there is also a high proportion of migrants. Gender, age and origin work are used as matching signals. But, because they are also discriminatory criteria they are neither displayed nor remunerated. That is why we can talk about “non-qualified” staff in the literal sense: its qualities are not valued nor recognized.
Schneider/Ali: To what extent do you think the theoretical perspectives within EC are useful in facilitating a smooth functioning and progress of society/societal institutions?
Marchal: The conventionalist approach emphasizes the need for agreement to coordinate, and the existence of alternatives: there are several “polities” (in French cités), several registers of action and valuation among which it is possible to choose. The driving force of such a society and what allows it to progress, are the controversies driven by the critical capacities of the actors. They reject unfair situations and attempt to make compromises between competing conventions. Of course, this is a utopian vision of social relations and the way institutions work. To be convinced, it is enough to take up the critical reactions that were made after the publication of the book De la justification: what place is given to power relations in such a society? De la critique, published by Luc Boltanski in 2009 (engl. 2011), aims to correct the situation by insisting on the mechanisms of oppression. Personally, I think that the pragmatic sociology of EC allows us to formulate good questions by looking at the conditions under which a company, a market, an activity (such as recruitment), could operate in an equitable way. Do institutions give real critical capacities to employees, consumers, applicants? Are these capacities well distributed among the layers of society? Does everyone have enough capabilities (using Amartya Sen’s vocabulary) to make choices? Through these questions, it is the place of democracy and liberalism that is being asked. This involves highly political choices …
Bessy, C., and Marchal, E., 2009. Le rôle des réseaux et du marché dans les recrutements: enquête auprès des entreprises. Revue française de Socio-économie, n°3, p.121-146
Boltanski, L and Thévenot, L., 1991, De la justification. Les économies de la grandeur. Paris: Gallimard.
Boltanski, L and Thévenot, L. 2006, On justification. Economies of Worth (Translated by Catherin Porter). Princeton: Princeton University Press.
Boltanski, L., 2009, De la critique. Précis de sociologie de l’émancipation. Paris: Gallimard.
Boltanski, L., 2011, On Critique : a sociology of Emancipation., London: Polity Press.
Chateauraynaud, F. 2003, Prospéro : Une technologie littéraire pour les sciences humaines, Paris: EHESS.
Desrosières, A. and Thévenot, L., 1988, Les catégories socioprofessionnelles. Paris: La Découverte.
Dodier, N., 1993. Les appuis conventionnels de l’action. Eléments de pragmatique sociologique, Réseaux Communication, Technologie, Société, n°62, pp. 63-85
Eymard-Duvernay, F. 2006. Pouvoir d’évaluation de la qualité du travail et décisions d’emploi. In Petit, Héloise and Thévenot, Nadine (eds.), Les nouvelles frontières du travail subordonné: Approche pluridisciplinaire (pp. 71-86). Paris: La Découverte.
Eymard-Duvernay, F. and Marchal, E., 1997, Façons de recruter. Le jugement des compétences sur le marché du travail. Paris: Métailié.
Ghirardello, A., 2005. De l’évaluation des compétences à la discrimination : une analyse conventionnaliste des pratiques de recrutement. Revue de Gestion des Ressources Humaines, n°56, pp. 36-48.
Granovetter, M., 1974. Getting a job: A study of contacts and careers, Cambridge: Harvard University Press.
Karpik, L., 2010. Valuing the unique. The economics of singularities. Princeton: Princeton University Press.
Larquier, G. de and Marchal, E. 2016. Does the Formalization of Practices Enhance Equal Hiring Opportunities? An Analysis of a French National Wide Employer Survey. Socio-Economic Review, p. 1-23.
Lendaro, A. and Imdorf, C. 2012. The use of ethnicity in recruiting domestic labour: A case study of French placement agencies in the care sector. Employee Relations, Vol. 34 Issue: 6, pp. 613-627.
Marchal, E. 2015. Les embarras des recruteurs. Enquêtes sur le marché du travail, Paris: EHESS.
Marchal, E. 2013. Uncertainties Regarding Applicant Quality. The Anonymous Resume Put to the Test. In Beckert, Jens and Musselin, Christine (Ed.), Constructing Quality. The classification of Goods in Markets (pp. 103-125). Oxford: Oxford University Press.
Marchal, E., Mellet, K. and Rieucau, G. 2007. Job board toolkits. Internet matchmaking and changes in job advertisements. Human Relations, n°60, pp. 1091-1113.
Schütz, A., 1987. Le chercheur et le quotidien. Phénoménologie des Sciences sociales, Méridiens Klincksieck.
Thévenot, L. 1984. Rules and implements. Investment in forms. Social Science Information, 23, pp. 1-45.