Some comments on ESEC outline of the talk

1
Some comments on ESEC outline of the talk

• Theoretical aspects, objectives what we agree
  on, what we suggest
• Measurement issues
• Coding ISCO
• Additional information
• Proposals
• More specific comments on matrix

2
Theoretical aspects, objectives

• We agree to construct use a european
  socioeconomic classification
• Main argument allow cross country comparability
• To achieve this, ESEC must be based on explicit
  foundations
• Allow cross country comparability
• Bonus clear interpretation in terms of own
  effect
• Not many theoretical frameworks available
• We agree on schema à la Golthorpe
• Causal interpretation of ESEC is in terms of
  employment relations

3
Theoretical aspects, main proposals

• On the ranking of ESEC groups
• independents employees have fundamentally
  different employment relations
• renumber groups 4 5 into 1 2
• What about life employment type contracts ?
• Current ESEC solely based on service
  relationship vs labour contract
• In France, life employment is an omitted
  important type of employment relation (full
  security contract)

4
About life employment

• Proposal add full security contract as form
  of employment regulation
• At the 2 digit level of ESEC ?
• 2 digit codes empty in countries where life
  employment does not exist
• France over 20 of employees have full
  security contracts high explanatory power

5
(No Transcript)
6
Measurement issues

• Coding 3-digit ISCO using french sources requires
  
• 4 digit occupation 4 digit activity
• In turn, this requires a large amount of
  information, not available in many sources
  relevant for ESEC validation (eg health,
  cultural participation)

7
Measurement issues

• Using the matrix requires additional information
  
• status of employment (independent / employee)
  number of employees if independent mostly OK,
  only issue is farmers
• Employees problem of identifying managers
  supervisors
• Within managers professionals, problem of
  identifying higher lower

8
Managers

• Use of firm size, not viewed as acceptable to
  distinguish higher from lower managers
• Constructing the category in French sources use
  of self-declared classification of employment in
  the firm (conventions collectives)
• Higher managers directors or one of their close
  assistants
• Lower managers all other executives (cadres)
  with commercial or administrative function
• Choice consistent with theory (higher
  senior)

9
Professionals

• A professional can be a manager in the above
  sense (eg head of RD)
• If so, should be treated as a manager
• Higher lower professional (based on level of
  expertise) can not often be distinguished
  (engineers, senior vs junior)
• In French sources, we use 4-digit occupation
  most engineers do not qualify as higher
  professional

10
Supervisors

• Missing variable cannot solely be inferred from
  occupation not usually available
• Add a new variable ? Also useful to define
  managers higher professionals (eg if
  engineer, number of persons under direct
  supervision)
• Otherwise acceptable approximation
  self-declared employment classification (agent
  de maîtrise)

11
Routine vs semi-routine

• Neither observed, nor precisely defined
• Major measurement issue
• French attempts to proxy by skill (firm
  classification of skills based on technical level
  of expertise, but also responsibility)
• Matrix V2 routine vs lower technical

12
More specific comments on matrix V1

• Independents no comment

13
More specific comments on matrix V1

• Employees Upper part of classification
• French approach leads to more elitist ESEC,
  consistently with theoretical definition
• gt Small group 1 directors, close associates
  specialized doctors, researchers, University
  teachers similar occupations group 2 reserved
  for plain engineers executives, teachers
  nurses, librarians belong to group 3, unless
  supervisor
• Matrix all executives more or less presumed to
  be higher managers most professionals presumed
  higher
• gtConsequence all in group 1 group 2
  reserved for nurses and the like

14
More specific comments on matrix V1

• Examples Upper part of classification
• Most professional (ISCO 21, 22, 24) gt group 1
  only exception teachers (ISCO 23) gt group 2
• Criticism a majority of professionals
  theoretically belongs to group 2
• All managers (ISCO 12) gt group 1
• Criticism most of them are not higher managers
  (plain administrative or commercial executives,
  even (or more so ?) in large firms

15
More specific comments on matrix V1

• Examples Upper part of classification
• Most administrative or sales associate
  professionals (ISCO 34) gt group 2
• Criticism no reliable information on
  supervisory powers (higher supervisory)
• Most clerks (ISCO 41, 42, including plain
  secretaries, supermarket cashiers) gt group 3
• Criticism most of them do not have sufficient
  expertise they would be better classified in
  group 7, unless additional information available

16
More specific comments on matrix V1

• Employees lower part of classification
• By default, french approach of ESEC also more
  elitist
• gt group 3 nurses, sales representatives,
  technicians group 6 supervisors upper
  fringe of production workers group 7 most
  secretaries, skilled employees group 8
  unskilled employees
• Matrix as nurses the like in group 2, most
  secretaries in group 3
• Difference stems from choices made in upper part
  of classification

17
More specific comments on matrix V1

• Example salespersons
• Matrix ISCO 52 gt group 7 ISCO 91 gt group 8
  
• Problem in France few persons coded in ISCO 91
• Solution use french PCS instead, with
  distinction between skilled unskilled
  salespersons

18
Conclusions

• Using matrix induces 2 sources of errors,
• at the stage of coding ISCO, in rows (eg France
  does not use the 10 employees threshold),
• At the stage of defining additional variables, in
  columns
• Countries have access to different information on
  occupation additional variables
• Matrix different across countries desirable,
  based on national classification of occupation
  all relevant information available

19
Conclusions

• Therefore ESEC consistent with theoretical
  foundations permitting cross country
  comparability is best achieved by providing
  guidelines based on ISCO rather than a
  ready-to-use matrix
• Validation in terms of consistency with
  theoretical principles to be carried out on this
  basis