Cet ouvrage fait partie de la bibliothèque YouScribe
Obtenez un accès à la bibliothèque pour le lire en ligne
En savoir plus

The International Standard Classification of Occupations (ISCO) in the European Union

5 pages
The revision of the International Standard Classification of Occupations (ISCO) offers the European Union an opportunity to harmonize its employment statistics. Despite their efforts, however, Eurostat and EU Member States failed to define a satisfactory common position during the International Labor Office’s 2004-2007 review. A new field of study is now opening up, with the aim of adopting the new version of the classification (ISCO-08) as a benchmark for the EU. The most promising approach is a judicious linkage between ISCO-08 and the EU’s ESeC project.
Voir plus Voir moins

The International Standard Classification
of Occupations (ISCO) in the European
! Jérémie Torterat*
The revision of the International Standard Classification of Occupations (ISCO) offers the European Union an opportunity
to harmonize its employment statistics. Despite their efforts, however, Eurostat and EU Member States failed to define
a satisfactory common position during the International Labor Office’s 2004-2007 review. A new field of study is now
opening up, with the aim of adopting the new version of the classification (ISCO-08) as a benchmark for the EU. The
most promising approach is a judicious linkage between ISCO-08 and the EU’s ESeC project.
he International Standard – informal-sector and self-Tc lassification of o ccupations consumption occupations such as
(ISco ) came into widespread use in production of goods for one’s own
the european Union (eU) in the 1990s household—activities often performed
(see article by Brousse) but has been by women—and subsistence
adapted in various ways. agriculture.
Its official eU adaptation, ISco - Another goal was to link the ISco
88(co M), is being used by eurostat, of updating to that of the International
course, but also by certain countries Standard Industry c lassification
such as Finland and Sweden. o ther (ISIc : see articles on classifications of
countries use specific national activities in this issue).
versions. Two examples are Spain
For this purpose, it was decided to and Portugal.
European Union
depart from the standard revision
The ISco revision offered an procedure, whereby the classification
planned for 2010. The Ic LS decision opportunity to promote convergence must be adopted successively by
came in response to a request among eU countries and to genuinely an Ic LS and by the Governing
from the United n ations Statistical harmonize employment statistics— Body of the ILo . The next Ic LS was
c ommission (Un Sc ) in March 2003. within a classification framework that scheduled for 2008, and the new
At the Un Sc ’s 35th session, in March eurostat did not intend, this time, to version had to be ready by end-2007
2004, the ILo pledged to complete adapt to eU practices. in order to be used in the upcoming
the updating by end-2007 in order census round. The adoption of the
to present the classification to the The eU countries, not the Union itself, new ISco was therefore assigned to
Un Sc in March 2008.are members of the International a tripartite meeting of labor-statistics
experts representing governments, Labor o ffice. eurostat’s role therefore
The need for the updating arose from employees, and employers.consisted in initiating consultations
the significant changes in the world of between countries to reveal
labor since ISco -88 was developed The 17th ILc S had recommended a convergence in certain areas and
in the first half of the 1980s. o ne limited updating, without challenging encourage countries to take common
area requiring examination consisted positions.
of occupations in information and
* Jérémie Torterat is Head of the Inter-regional communication technologies (Ic Ts),
Resources c enter for Household Surveys in
General organization as most of them were new and thus French overseas départements (Do Ms), at
In See ’s Regional o ffice in La Réunion. He was of the process very poorly classified in ISco -88.
previously an analyst at the Jobs and Skills
Department of the Directorate for Research,
In December 2003, the 17th Two other areas were also singled out Studies, and Statistics (DAReS), a joint unit of
the French ministries in charge of labor and International c onference of Labor for special attention:
Statisticians (Ic LS) called for an ISco 1. o riginally published as “La classification
internationale type des professions dans l’Union updating. The revised classification – occupations in which women are a
européenne,” Courrier des statistiques (French
would meet the needs of the next majority, for example by disaggregating
series), no. 125, n ov.-Dec. 2008, pp. 22-26,
world population-census round the “secretaries” group http://www.insee.fr/fr/ffc/docs_ffc/cs125e.pdf.
Source: WikipédiaThe International Standard classification of occupations (ISco) in the european Union
ISco principles and concepts. (French: emploi) than to the French the same classification level, whereas
Moreover, a major updating—such concept of profession without being their tasks are not very specialized
as the shift from ISco-68 to ISco- totally equivalent. ISco defines these and should therefore not be placed in
88—was a difficult proposition to concepts (Box 2), but maintains an residual categories.
contemplate for many countries using ambiguity about the entities that it
an ISco-88 model. organizes. In the French version, it A final measure envisaged in the
theoretically classifies jobs (emplois), updating was to set up groups for
Accordingly, no changes were but actually does so by occupation “professional civil servants,” whose
planned at the higher level of the (profession). For this purpose, ISco occupations have no equivalents in
classification. Besides classifying new contains a set of classification priority the private sector.
occupations, the updating would aim rules. When the tasks relating to a job
instead to merge, split, and reposition concern more than one unit group, Two successive questionnaires were
ISco unit groups and rewrite certain the job is classified in the category sent at end-2004 and in mid-2006 to:
descriptions. that represents the highest skill level.
When the same skill level comprises – national institutions concerned,
The ILo Bureau of Statistics, which several tasks, the job is classified in including statistical institutes, labor
was masterminding the updating, had the unit group corresponding to the ministries, and employment agencies
to engage in consultations both on task that requires the largest share of (State-run exchanges)
the general concepts and on specific working time. This arrangement, which
occupational groups on the basis of is, theoretically, of narrow import, – employers’ and employees’
IcLS recommendations. has major practical consequences organizations
for ISco-based comparisons. For
In particular, the ILo was tasked with – specific international organizations,
surveying the national applications of such as o MS for health-related
the “skill level” criterion in countries occupations and FAo for agriculture-
that use it to classify occupations. related occupations.
Special attention would be paid to
(1) the choice of titles, in order to Many trade associations had appealed
minimize the risk of misinterpretation, directly to the ILo, asking that the
and (2) “not elsewhere classified” occupations they represented should
(n.e.c.) categories, also called residual be classified separately in order to
categories. give them a higher profile and greater
weight in the countries that will adopt
In sum, the updating would draw ISco.
lessons from the experience gained
from ISco -88 implementation in
International Labor Office
many countries. The EU in the updating
Indeed, successive versions of example, it poses a problem for the
ISco have served as classification In the wake of the decision to update establishment of the eU register of
“standards.” Application problems 2 ISco , eurostat and the european occupations, euroccupations. one
were often common to several national statistical institutes (nSIs) of the project’s assumptions is that a
countries. o ne example was the were strongly encouraged to supply classification in the same unit group
proliferation of parallel groups, which a coordinated response to ILo. This for different countries necessarily
describe occupations that are similar approach paralleled the launch by the implies unquestionable comparability
but linked to a very diverse range of european commission and eurostat between the occupations studied: the
skill and education levels, according of a project to develop a european “same occupation” (même métier)
to specific national legislation. socio-economic classification (eSec: should—owing to a confusion between
This profusion raised major data- see article by Brousse).job (“occupation”) and métier—imply
comparability problems: some comparable working conditions or
concepts such as “manager” were To facilitate eurostat coordination pay.
not even defined. and allow in-depth discussion, four
We may, however, consider that regional meetings were held between
Lastly, the ISco revision could offer the ISco-08—like its predecessor ISco- June and December 2005. They
opportunity to minimize a weakness 88(co M)—classifies occupations covered ISco updating, the eSec
of the classification as an international rather than jobs. In this approach, the project, and the links between the two.
standard: “occupation” in the english question is whether it is useful to have ILo representatives were invited.
sense is closer to the concept of “job” general categories of occupations, i.e.,
sets of jobs whose tasks are common Statistics n orway hosted
2. www.euroccupations.org. to several occupational groups at representatives from the Baltic and
courrier des statistiques, english series no. 15, 2009 23
Source: ILoJérémie Torterat
Scandinavian countries. Portugal’s Ine the number of employees. However,
convened delegates from Belgium, this simple criterion was not always
France, Italy, and Spain,. The Greek viewed as capable of ensuring
n SSG invited colleagues from c yprus, adequate comparability between the
Germany, and Switzerland. Hungary’s different types of “managers.” The
Hc So received statisticians from problem is that the links between the
Bulgaria, the c zech Republic, Ireland, number of employees in the local unit
the n etherlands, Romania, Slovakia, or firm and the management structure
and Slovenia. Participants were are not always clear.
asked to focus their presentations on
national positions regarding several For the rapporteurs, the use of the
questions in the questionnaire sent by number of employees in the local
Eurostat, entrance to Bech Building, Luxembourg
the ILo to n SIs in December 2004. unit was not satisfactory. They
suggested modifying the definition
The proceedings were summarized of the “managers and managers of introduction of a new minor group
by Peter elias and Margaret Birsh small enterprises” to bring it closer in the latter for “public[-]service
(Institute for employment Research, to the notion of “owners of their administrative professionals.” The
University of Warwick), ISco - own small enterprises.” While such a rapporteurs noted a consensus in
updating consultants for eurostat and shift would make the definition more favor of retaining the category and
status-determined, the rapporteurs promoting it in the updating. However,
believe that the specificity of the the available summary of n SIs’
tasks performed by such managers, specific comments on this point do
compared with those of a manager not clearly indicate such a conclusion.
working in a large hierarchical structure, The ILo , for its part, came out against
justify their separate classification. the inclusion of this specific category
at minor-group level in the new ISco ;
Another important point discussed instead, the ILo proposed introducing
was the identification of “supervisor” a unit group of “administrative
occupations, as “supervision” is one professionals” in a minor group of
of the main criteria used to construct “business professionals”—a term
eSec classes. ISIc -68 frequently whose translation into French is a
identified these occupations and challenging task. The rapporteurs
classified them separately in a specific endorsed the ILo proposal. They
sub-major group. ISIc -88 did not stipulated that the accompanying
preserve this distinction. According definitional notes should clarify
to the ILo Bureau of Statistics, the that this group consists of those
decision was due to two factors: occupations which fall primarily (but
not exclusively) within the public
– data-collection problems: in most sector and for which a high-level
FAO headquarters, Rome
cases, a special question is required education is a requirement.
in order to identify supervisors
promoters of the eSec project. We eurostat and the european
reproduce the main conclusions of their – information loss due to the c ommission’s Directorate-General
abstract, published in The Review of classification in the same unit group for Research (DG-Research) strongly
ISCO88: a European perspective. The of occupations requiring a very wide advocated the inclusion of categories
document emphasizes the common variety of technical skills. specific to researchers. However,
difficulties connected with managerial most eU Member States declined
o ne possible solution, adopted and supervisory occupations— to accept the proposal. Few of them
by some countries, is to identify “managers” and “foremen/women identify research occupations in
“supervisors” thanks to an additional and supervisors”—and government their national classifications or wish
level of the classification, but the occupations. to do so. France, which identifies
rapporteurs prefer the introduction researchers in the public sector only,
The n SIs that participated in the of certain categories of “supervisors” took a distinct position.
regional meetings consistently in sectors where “supervision” is
underlined the difficulty of classifying The final issue raised was the recognized as a specific set of tasks
“managers” according to the number classification of “wholesalers/that constitute the majority of
of other “managers” employed in the retailers,” which re-emerged as a within the job.
organization (criterion proposed by critical point at the tripartite meeting.
ILo for ISco -88). ISco -88(co M) o ne of the main differences between Their classification in major group
suggested proxying the criterion by ISco -88 and ISco -88(co M) is the 1 (“managers”) may seem the most
Source: Wikipedia
Source: eurostatThe International Standard classification of occupations (ISco) in the european Union
logical answer, but it creates sizable consensus would emerge on various
heterogeneity in the major group. points. Box 1: “Supervision” in Europe
Spain adopted the solution in cno-94,
According to the Warwick IeR
and the rapporteurs recommended it Absent a classification well suited
rapporteurs, Peter elias and
at the conclusion of these meetings. to the eU, the Union’s n SIs and Margaret Birsh, a consensus
eurostat must now work to ensure the emerged over the need to identify
convergence of coding procedures, “supervisors” separately.
A temporary failure so as to finally obtain a corpus of truly If we review the points raised by
harmonized and usable data. It makes representatives of the Member
In theory, the meetings were an sense, therefore, that ISco coding by States in the summary, we find that
the situation is complex.excellent initiative. While not forming the Member States should have been
an image of the world in miniature, Some countries are unreservedly in the topic of the first nSI meeting to
favor (Bulgaria, Slovakia, and Spain, the eU comprises countries with assess the european Socio-economic
as well as cyprus and Slovenia for extremely diverse economic and classification (eSec) project—whose
certain specific areas). others are social structures. The exchanges standard implementation should be
equally in favor but raise major
could foreshadow those that would based on ISco , according to the problems relating to data collection
take place during the adoption of the recommendations. We lack an eU or identification via occupation
classification. They should, therefore, classification, but we could seek a descriptions (czech Republic,
have interested the ILo as much as common interpretation of the new Germany, and Hungary).
eurostat. ISco. Some countries take no position
and make no comments (Belgium,
The meetings might have devised, Finland, Latvia, netherlands,
Portugal, and Romania). others, for example, a consensual solution Toward an EU
without pronouncing themselves among n SIs for the classification interpretation of ISCO-08?
on the fundamental issue, mention of wholesalers/retailers that would
data-collection problems (Denmark,
have satisfied the netherlands (where There are several standard alternatives estonia, Greece, n orway, and
most retail sales take place in large to achieve convergence of eU-country Sweden).
stores) as much as France or Spain codes in the new ISco. As regards classification, Italy
(which have large networks of small and Lithuania prefer to include
Creating an additional level in ISCO retailers). The europeans could have “supervisors” with persons
would make it possible to define “supervised.” France notes that offered such a solution to the ILo,
only certain jobs can be identified. specific categories without being with a significant chance of adoption
Switzerland favors identification obliged to describe all the occupations by the other States. The emergence
using an ancillary variable.in extreme detail. This is the type of of such a consensus at eU level might
have allowed the ILo, at a relatively obstacle that revealed the limitations
early phase of the process, to make of ISco-68, which identified more
be used for the same job in the same
relevant proposals. But the meetings than 1,500 occupations. Many
language, but in different countries:
ended in failure. Why? occupations, for instance, could be
a female kindergarten teacher is an
usefully defined with information on
institutrice de maternelle in France and
First, holding the meetings by the activity sector. This could apply
a jardinière d'enfants in Switzerland.
region generated a weakness: the to management personnel, of course,
occupational structures and tasks in but also to “non-managing managers” The use of ancillary variables,
the countries attending the meetings (such as road-traffic engineers) or to supplement individuals’ self-
are often similar, so that some basic even to “non-managers” such as descriptions, seems necessary to
divergences did not emerge, despite secretaries. The distinction between obtain relevant codings. They include
the organizers’ efforts. Perhaps, public and private sectors, which tasks performed, occupational
in order to succeed, the meetings ISco does not apply, would become position, and education level.
should not have taken the form possible.
of consultations. Because of this A more interesting and original
format, the nSIs had no incentive to Another path to progress is the approach would be to conduct eSec
recognize the european dimension enhancement and harmonization and ISco-08 coding harmonization
of certain issues. Ultimately, it was of the alphabetical index of as part of a single conceptual
eurostat alone that argued for what occupations. The standard ILo index, process, since the construction of
the EU position should be. This while available in english, French, eSec classes relies on ISco coding
sent each State back to its national and Spanish, cannot solve all coding and a few ancillary variables. Ideally,
issues. By placing itself in the same issues. The identification and naming the key criteria of both classifications
position as the nSIs, eurostat of occupations are rooted in very should be homogeneous.
probably sacrificed its potential role diverse sociolinguistic contexts that
as organizer of eU statistics and, need to be taken into account. For The prospect of reallocating
paradoxically, the possibility that eU exemple, totally different names can occupational unit groups makes it
courrier des statistiques, english series no. 15, 2009 25Jérémie Torterat
possible to examine, for each group Very generally, the prospect of
and on a country-by-country basis, reallocating occupations raises Box 2: Jobs, skills, tasks,
duties, specializations, and the validity of the classification in the question of determining what
5occupationsterms of the employment relationship exactly we want to measure by
recommended in the eSec theoretical reaggregating subsets into a socio-
The resolution offers the following
framework. That is the subject, for economic classification. As in all definitions:
taxonomic problems, the goal is the example, of an article by Michael – ISco -08 classifies jobs; a
3 initial question that should determine Tåhlin, which is critical of esec. job is defined as a set of tasks
the theoretical choices. The choices and duties performed, or meant
Tåhlin’s paradoxical conclusion is may even be arbitrary, as long as to be performed, by one person,
that the theoretical underpinning of they make sense. That is the meaning including for an employer or in
self[-]employment.the eSec project—most notably, the of Tåhlin’s paradoxical conclusion:
concept of employment relationship— – An occupation is defined as a we should use esec classes while
set of jobs whose main tasks and should be abandoned, but not reviewing their theoretical framework
duties are characterized by a high necessarily the classes proposed a posteriori in order to redirect it
degree of similarity. A person may by the project. Tåhlin suggests toward the explanatory dimensions
be associated with an occupation
reinterpreting the classes in terms that we can and want to display. through the main job currently held,
4of education level, and to reassign a second job or a job previously
certain “occupations x education” We should therefore try to use the held.
cross-tabulations to strengthen the suggested classes in different spheres, – Jobs are classified by occupation
classes. If that were done, the use of so as to identify the social areas where with respect to the type of work
ISco and ancillary variables would not they are interpretable and where performed, or to be performed. The
be an artifact to proxy an evanescent they give meaning to an analysis. basic criteria used to define the
system of major, sub-major, minor, theory of employment relationships. We can envisage other criteria than
and unit groups are the “skill level” education, or even reconsider certain
and “skill specialization” required to The education criterion is not always thresholds in the additional variables
competently perform the tasks and
easy to use, as national legislations propose by eSec. This integration of
duties of the occupations.
may sometimes differ significantly. eSec constraints into ISco coding
The employment relationship is far standards could also allow a
from being stabler in time and is of socio-economic classes from a
even harder to quantify. Moreover, higher level of the classification and
a “craft” or “trade” context (self-
the “skill level”—a similar concept—is using fewer additional variables,
employed mechanic, or butcher,
one of the explicit ISco classification which would result in the use of
whose distinctive status lies between
criteria. As the eU is a set of countries these classes in a larger number of
employee and employer) might thus
with more consistent standards of sources.
be settled at eU level and allow data
formal education and on-the-job
harmonization. n
training than ILo countries, one could In particular, we could consider using
try to express “skill levels” in the the eSec criteria of enterprise (or local-
corresponding eU classifications. unit) size (currently 10 employees,
but thresholds of 5 or 50 employees
This linkage between eU classifications would be relevant as well), or even
is desirable in itself, but even more so employees’ statuses by assessing the
3. M. Tåhlin, “class clues,” European
in the context of the eSec project, if corresponding variables contained in Sociological Review, vol 23, no. 5, 2007.
4. In this context, not restricted to formal it emerged that education could be a eU sources, in order to standardize
education, but extending to on-the-job training.
workable criterion for defining socio- choices between categories. The 5. excerpt from resolution adopted by ILo in
economic classes. coding of occupations performed in December 2007.