Comparing Job Quality Across Countries & Over Time Ekaterina - - PowerPoint PPT Presentation

This project has received funding from the European Union’s Seventh Programme for Research, Technological Development and Demonstration under Grant Agreement No 312691

Comparing Job Quality Across Countries & Over Time

Ekaterina Kalugina University of Evry and Center on Employment Studies Ekaterina.Kalugina@univ-evry.fr

www.inclusivegrowth.be www.inclusivegrowth.be 2

Overview

• 1. Operationalizing the concept of quality of

working life

• 2. Comparing quality of working life across

countries

• 3. Comparing quality of working life over time
• 4. Empirical cross-country studies of change in

quality of working life using EWCS

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 3

Introduction

• « One major obstacle to foster concrete policy

actions on job quality has been the challenge of defining and measuring job quality in ways that are amenable to comparisons over time and across countries and socio-demographic groups » (OECD, 2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 4

Introduction

• « Another obstacle is that job quality is inherently a

multi-dimensional concept that can be measured in many different ways, and which does not have the same meaning for different individuals in different countries » (OECD, 2015)

• “Comparability is ensured when concepts and

definitions follow internationally agreed standards and the surveys/instruments from which data are collected are based on a harmonised questionnaire and similar implementation design.

• However, comparability can also be achieved by putting

together broadly comparable instruments ex post” (OECD, 2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 5

Operationalizing the concept of quality of working life

• Comparisons over time and across countries:

– Defining quality of work:

• Creation of indicator(s) of quality of work that can be used to

compare job quality across countries and over time

– Objective measures versus subjective measures – Data collection

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 6

Operationalizing the concept of quality of work

Objectivist versus Subjectivist approach

• Objectivist approach

– Objective indicators

• Job attributes that can be observed by a third party (OECD,

2015)

– Number of fatal and serious accidents; Number of days lost due to sickness (UNECE – Business Europe, 2001)

– Individuals’ self-assessment of their own situations

• Green et al. (2013): “Objective concept of job quality: a set of

features that help to meet jobholders’ needs from work. This approach means that we include only variables characterizing jobs, leaving out those applying to individuals’ lives or their preferences.”

• Workplace relationships, for example

Title/date

www.inclusivegrowth.be www.inclusivegrowth.be 7

Operationalizing the concept of quality of work

Objectivist approach

• Multidimensional concept
• Choice of relevant dimensions

– Theoretical and conceptual considerations

• For work organization

– Job demands-job resources theory (Demerouti et al. 2013) Resources: workplace relationships, work autonomy Demands: time pressure, physical health risk factors, workload – Job demands – job controls model (Karasek, 1979; Karasek and Theorell, 1990) Job demands: psychological demands (work load, work intensity), Job controls: Job discretion, social support , learning opportunities

• Sen capability concept (Sen, 1985)

– Functionnings: “are beings and doings that people value and have reason to value” – Capabilities: “are a set of vectors of functionings, reflecting the person’s freedom to lead

• ne type of life or another (…) to choose from possible livings”

– Could include subjective measures – This approach doesn’t provide the list or the hierarchy of relevant dimensions

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 8

Operationalizing the concept of quality of work

Objectivist approach

– Example of OECD (and not only OECD) response

– “Focusing on those aspects of a job that have been shown to be particularly important for workers’ well-being” (also Green et al. 2013; Clark, 2015) – Focus on individual workers

– For each dimension to choose a limited set of indicators (comparable across countries and over time) – Simplicity of indicators is privileged for better comparability across countries and over time

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 9

Operationalizing the concept of quality of work

Objectivist approach

• Aggregation problem within dimensions and between

dimensions

– If a dimension contains multiple items: how to aggregate them? – Overall index of quality of work: how to aggregate different dimensions? – The weighting structure

• Equal weighting and/or averaging
• Use of more sophisticated statistical tools

– PCA : Principal Component Analysis – MCA : Multiple Correspondence Analysis – More recently: Fuzzy logic

• Use of satisfaction equation

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 10

Operationalizing the concept of quality of work

• Subjectivist approach

– Purely subjective measures

• Job satisfaction questions or satisfaction of different aspects of

working life (working conditions, for example)

• Objectivist approach is considered to be better

for the comparability across countries and over time (OECD, 2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 11

Operationalizing the concept of quality of work

• The data

– OECD Inventory: stats.oecd.org/Index.aspx?DataSetCode=JOBQ

• Three main international sources that are currently available

– EWCS (European Working Conditions Survey) of Eurofound: http://eurofound.europa.eu/surveys/ewcs/index 1991, 1996, 2000/2001, 2005, 2010, 2015* EU 28, Norway, Turkey, FYROM, Albania, Kosovo and Montenegro – International Social Survey Programme (ISSP Work Orientations Module, 2005): http://www.gesis.org/issp/home 1989, 1997, 2005, 2015/2016 Numerous countries – European Social Survey (ESS): http://www.europeansocialsurvey.org 2004, 2010 EU28, Norway, Switzerland, Turkey, Israel, Russian Federation, Ukraine

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 12

Comparing Quality of Working Life across Countries

How to uncover the “country effects” ? I.e. to identify how different policy environments and institutions affect socio- economic outcomes? Institutional theories

• Varieties of Capitalism (Production Regime Theory):

Soskice (1999), Hall and Soskice (2001) Crucial difference between societies is the level of employer coordination: coordinated market economies (Germany, the Scandinavian countries) versus liberal market economies (Britain, US and Ireland)

• Power Resource (Employment Regime Theory):

Korpi (1983 and 2006), Esping-Andersen (1990), Gallie (2007) Crucial difference between societies is the relative power resources of employers and employees: distinction between inclusive systems (Nordic) of employment regulation, dualist systems (Continental) and liberal systems (UK, Ireland)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 13

Comparing Quality of Working Life across Countries

How to uncover the “country effects” ? I.e. to identify how different policy environments and institutions affect socio- economic outcomes?

Empirical (regression based) methods

• If country effects exist it induces correlations across
• bservations within the countries: the standard errors (SE)

estimates are downwardly biased

• Pooling the data for all countries and using cluster-robust standard

errors:

This approach just controls for the multilevel (individual + country) structure of the data by using country-cluster-robust SEs but does not uncover the country effect

• Making analysis by country
• Create dummy variable for each country or group of countries
• Multilevel analysis
• Other methods (Bryan&Jenkins, 2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 14

Comparing Quality of Working Life across Countries

Basic linear model:

with ;

– outcome variable (quality of working life index, for example) for each person – variables that summarise individual-level characteristics (age, education etc.) –

variables that summarise country-level features (for

example, unemployment rate, trade-union coverage) – - country effects – - individual effects – and are each assumed to be normally distributed and uncorrelated with and

• Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 15

Comparing Quality of Working Life across Countries

• Empirical methods
• Separate models for each country

– The country effect is absorbed into the intercept term in each country’s regression model and cannot be identified separately from – Every model parameter is country-specific – The country effects are treated as fixed parameters

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 16

Comparing Quality of Working Life across Countries

• Empirical methods
• Country fixed effects (FE) model

– Pooled data but the model includes distinct country intercepts (estimated as the coefficients on country binary indicator variables) – Each country intercept represents the effects of unobserved factors that are shared within each country – The individual-level characteristics could be also allowed to differ between countries by including the interaction terms

• f individual-level characteristics with the country indicator

variables – Country-level variables cannot be included as additional predictors

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 17

• Green and McIntosh

(2001), « The intensification of work in Europe », Labour Economics ,8, p.299

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 18

• Green and McIntosh (2001), p.301

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 19

Comparing Quality of Working Life across Countries

• Empirical methods
• Country random effects (RE) models (Multilevel models)

– This approach allows for country effects which are modelled as random (with the normal distribution) – – Multilevel models are designed to analyze variables from different levels simultaneously, using a statistical model that properly includes the various dependencies (Hox, 2010) – The main advantage: country-level regressors could be used – This approach is the natural choice if the focus is on the effects of country-level predictors or the variance component structure – In simple case and are fixed while and are random – It is also possible to allow for variation in slope parameters

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 20

Comparing Quality of Working Life across Countries

• Empirical methods
• Country random effects (RE) models (Multilevel

models)

– – The key parameter to estimate is the intra-class correlation:

• is the variance of the country random effects
• is the variance of the individual random effects

It gives the expected correlation between two randomly chosen units that are in the same country and measures the share of the total variance that occurs between countries

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 21

Comparing Quality of Working Life across Countries

• Empirical methods
• Country random effects (RE) models (Multilevel

models)

– The popularity of regression analysis of multilevel country data is illustrated by the European Sociological Review (Bryan & Jenkins, 2015)

• Of the 340 articles published between 2005 and 2012,

approximately 75 exploit multilevel data sets with individual respondents within countries.

• Multilevel models are used in 43 of the 75 articles (57%; or 13%
• f all 340 articles)

– In other social science journals: 14 of the 111 articles published in the Journal of European Social Policy between 2005 and 2009

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 22

Comparing Quality of Working Life across Countries

• Empirical methods
• Country random effects (RE) models (Multilevel models)
• Main problem of this method: usually large sample sizes
• f individuals within each country but only a small number
• f countries available (Bryan & Jenkins, 2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 23

Comparing Quality of Working Life across Countries

• Empirical methods
• Country random effects (RE) models (Multilevel models)
• In this case it was argued that analysts can reliably estimate

individual-level effects but estimates of parameters summarizing country effects are likely to be unreliable

– Estimates of country-level fixed parameters are likely to be estimated imprecisely: users will conclude too often that a country effect exists when it does not – Only fixed parameters on individual-level predictors are unaffected by having a small number of countries

• How may countries does one need for multilevel model

analysis of multi-country data to provide reliable estimates?

– Last results show (Bryan&Jenkins, 2015): at least 25 countries for linear models and at least 30 countries for logit models

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 24

Comparing Job Quality Between Countries

• Other methods (Bryan & Jenkins, 2015)

– (1) Supplement regression-based modelling with more descriptive analysis of measured country differences.

• The two-step approach for example

– (2) Explore methods that are more robust when there are small numbers of countries

• Small sample corrections (in SAS or R), bootstrapping
• However some of these require specialised knowledge and are

available in only a few software packages.

– (3) Move beyond classical (frequentist) statistics and make greater use of Bayesian methods of estimation and inference

• The problem is that these methods also require statistical

expertise beyond that of most applied social science researchers, as well as specialist software.

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 25

Comparing Job Quality Between Countries

• Other methods (Bryan & Jenkins, 2015)

– Two-step approach

• One regression at the individual level
• a fixed effect for country

In practice, this could be obtained by letting be a country- specific binary indicator variable in an OLS regression

• Another regression at the country level
• is in our case the coefficient on the country indicator variable

Estimated by GLS (correcting for heteroscedasticity) or OLS if the number of countries is sufficiently large

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 26

Bryan&Jenkins (2015); Supplementary materials

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 27

Bryan&Jenkins (2015); Supplementary materials

Across countries & over time/10.05.2016

This figure illustrates the visualisation method of assessing country-level differences. It replaces step2 of the two-step method with graphs of the estimated country intercepts (from step 1) against childcare spending

www.inclusivegrowth.be www.inclusivegrowth.be 28

Comparing quality of working life over time

• Identifying trends in job quality

– Is there an upward or downward trend in quality of working life? – Is the trend significant?

• Explaining trends in job quality

– What are the reasons why the quality of working life is changing?

• Convergence or divergence

– Does the job quality converge or diverge between groups or countries?

• Divergence will occur when varying rates of growth exacerbate

differences in job quality

• Convergence will occur when different growth rates decrease

differences

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 29

Comparing quality of working life over time

• Would we expect quality of working life in Europe to

improve or decline?

– To improve: “An economy that was more knowledge intensive and involved increasingly complex technologies required more skilled employees. This in turn was seen as likely to lead to a long-term enhancement of the quality of work”, Gallie (2013) – To decline: persistence of neo-Fordist labor processes, globalisation (Green et al., 2013)

• Would be expect job quality in Europe to converge
• r diverge? (Holman et al., 2015; Green et al., 2013)

– To converge: “one best way”, diminishing institutional differences, EU legislation – To diverge: context-dependent strategies, national institutional differences

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 30

Comparing quality of working life over time

• Available data on quality of working life are mostly repeated

cross-sectional surveys

– There is no panel data available on individual level – The exception in Europe is EU-SILC (EU statistics on income and living conditions) data with longitudinal data pertaining to individual- level changes over time, observed periodically over a four-year period.

• However there are no questions about intrinsic work characteristics
• Empirical methods to uncover trends over time

– Mostly descriptive

• Comparing averages, percentages or more complicated synthetic indicators
• ver time
• Dummy’ indicators for time trend within the regression analysis
• Pseudo-panel data construction: the observation is the average score of

the cohort constructed around particular characteristics (for example: country, occupation, year)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 31

Comparing quality of working life over time

• Explaining trends in job quality

– Regression technics using factors at different levels of analysis (Greenan et al., 2014; Holm & Lorenz, 2015)

• Individual characteristics

– Gender, age…

• Intermediate-level factors

– Sector, type of work…

• Macro-economic factors and national institutional characteristics

– The varieties of capitalism framework (Production Regime Theory) Culture of employer coordination (bargaining coordination) – The employment regime framework (Power Resource) Strength of trade unions

– Latent growth modelling (LGM) (Holman et al., 2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 32

Comparing Job Quality between Countries and Over Time: some empirical studies

– Greenan, Kalugina & Walkowiak (2014), « Has the quality of working life improved in the EU-15 between 1995 and 2005? », Industrial and Corporate Change – Green, Mostafa, Parent-Thirion, Vermeylen & Van Houten (2013), « Is job quality becoming more unequal? », Industrial and Labour Relations Review – Holm & Lorenz (2015), « Has ‘discretionary learning’ declined during the Lisbon Agenda? », Industrial and Corporate Change – Holman, Holtgrewe, Rafferty, Ramioul & Vandekerskhove (2015), Convergence and divergence of job quality in Europe 1995–2010, Publications Office of the European Union, Luxembourg

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 33

Greenan et al. (2014)

• Aims

– Provide a mapping of quality of working life in EU-15 – Measure its evolution – Assess the sensitivity of observed trends to « composition effects » and « country effects »

• Time:

– 1995 – 2000-2005

• Data:

– EWCS

• Country coverage:

– EU-15

• Method

– Three dimensions of quality of working life (objectivist approach)

• Physical strain
• Work intensity: technical constraints/customer constraints
• Work complexity

– Weighting structure: Multiple Correspondence Analysis (MCA) – Descriptive analysis – Multilevel modelling

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 34

Greenan et al. (2014)

• Quality of physical working conditions

– Breathing in vapours, handling dangerous products; – Radiation, painful or tiring positions – Exposure to vibrations, noise, high/low temperatures; – Carrying or moving heavy loads – Repetitive hand or arm movements – Wearing personal protective equipment – Harassment and discrimination at work

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 35

Greenan et al. (2014)

• Work Intensity

– Repetitive tasks of less than 10 minutes – Working at very high speed – Working to tight deadlines – Pace of work dependent on:

• Work done by colleagues
• Direct demands from people
• Numerical production targets
• Automatic speed of a machine
• Direct control of boss

Across countries & over time/10.05.2016

UE-15

1995-2005 1995-2000 2000-2005

Intensity of technical constraints Intensity of market constraints

www.inclusivegrowth.be www.inclusivegrowth.be 36

Greenan et al. (2014)

• Work Complexity
• Main job involve:

– Meeting precise quality standards – Self-assessment of the quality of work – Problem solving – Monotonous tasks – Learning new things – Complex tasks – Rotating tasks between yourself and colleagues

• Possibility of choosing or changing the order of tasks and

methods of work

• You can get assistance from colleagues
• You can take your break when you wish
• You are free to decide when to take holidays or day off

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 37

Greenan et al. (2014)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 38

• The intercept gives the average level of each indicator for the reference individual in an

average EU-15 country in 1995

• Time trend

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 39 Across countries & over time/10.05.2016

Variables that are positively linked to the development of a knowledge-based economy are positively correlated with the degree of work complexity Countries with a larger part of educated people have a higher degree of work complexity

www.inclusivegrowth.be www.inclusivegrowth.be 40

Greenan et al. (2014)

• General conclusions

– Significant DECREASE in work complexity even after controlling for micro and macro characteristics – Significant INCREASE in intensity of technical constraints – Significant INCREASE of market constraints after controlling for micro and macro characteristics – Significant INCREASE in physical strain

• Country conclusions

– Problematic situation in Mediterranean countries especially Greece and Portugal – Deterioration of all quality of working life indicators in Germany and Italy

• « Complexity paradox »

– A negative residual effect in the evolution of complexity, measured with a multilevel model that takes into account characteristics of workers and countries – This effect is all the more important that structural evolutions (by age,

• ccupations, education or ICT development) should favour an increase

in complexity – What can explain this paradox ?

• « Objective » and « Subjective » explanations

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 41

Green et al. (2013)

• Aims

– Examine trends in nonwage aspects of job quality in EU-15 – Measure both the level and the dispersion of job quality between countries and between groups

• Time:

– 1995 – 2000-2005

• Data:

– EWCS

• Country coverage:

– EU-15

• Method

– Four dimensions of job quality are examined (objectivist approach)

• The Work Quality
• The Work Intensity
• The Good Physical Environment
• The Working Time Quality

– Weighting structure: averaging/summation/normalization (on scale 0-100), i.e. equal weighting – Descriptive analysis of the evolution of the average indicators and its dispersion over time

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 42

Green et al. (2013)

• Work quality

– Generally, does your main paid job involve:

• Solving unforeseen problems on your own?
• Complex tasks?
• Learning new things?
• Working with computers, PCs, network, or mainframe?

– Over the past 12 months, have you undergone any training paid for or provided by your employer, or by yourself if self-employed? – Are you able to choose or change:

• Your order of tasks?
• Your methods of work?
• Your speed or rate of work?

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 43

Green et al. (2013)

• The Work Intensity

– Does your job involve

• Working at very high speed?
• Working to tight deadlines?

– On the whole, is your pace of work dependent, or not,

• n:
• The work done by colleagues?
• Direct demands from people such as customers, passengers,

pupils, patients, etc.?

• Numerical production targets or performance targets?
• Automatic speed of a machine or movement of a product?
• The direct control of you boss?
• The Good Physical Environment

– Idem

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 44

Green et al. (2013)

• The Working Time Quality (WTQ)

– How many hours do you usually work per week in your main paid job? – How many times a month do you work at night, for at least 2 hours between 10:00 pm and 05:00 am? – How many times a month do you work in the evening, for at least 2 hours between 6:00 pm and 10:00pm? – How many times a month do you work on Saturdays? – How many times a month do you work on Sundays?

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 45

Green et al. (2013)

Test of validity of job quality indicators

Across countries & over time/10.05.2016

All variables are highly significant and impact the subjective well-being in expected way

www.inclusivegrowth.be www.inclusivegrowth.be 46

Green et al. (2013)

Job quality according to three of the indices is quit stable over time. A substantive rise of 5.4 points occurs over time in the Working Time Quality index, accompanied by a fall of 0.025 points in the Gini coefficient (inequality)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 47

Green et al. (2013)

The biggest increase in Work Quality is for those with temporary employment agency contract (crisis?). Work Quality increased for professionals while it decreased for service workers

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 48

Green et al. (2013)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 49

Green et al. (2013)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 50

Green et al. (2013)

Conclusions

• Global stability

– The levels and distributions of the job quality indices across individuals and groups remained relatively stable over the period – However a steady rise in the Working Time Quality index

• Countries analysis

– The relative stability reflects the size dominance of “big” countries: France, Germany and the UK – Improvement of Work Quality in some countries – Intensification of work in some countries

• Dispersion analysis

– The falls in the dispersion of Good Physical Environment, Working Time Quality and Work Intensity – A pattern of slow national convergence in the means and dispersion of these three indicators

• Socioeconomic groups analysis

– Mostly stable gaps in job quality between men and women – Rise in work intensity of temporary agency workers and – Decline in work quality among service workers

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 51

Holm & Lorenz (2015)

• Aims

– Explore both differences in work organization between European nations and changes within them – Investigate the institutional and economic conditions and changes that may account for the way work organization evolved within European nations

• No explicit reference on the quality of working life but the forms of work organization

contribute to both performance and individual well-being

• Background: the relation between work organization, employee learning and national

innovation performance in Europe

• Time:

– 2000-2010

• Data:

– EWCS

• Country coverage:

– EU-25

• Method

– 15 binary variables describing work content, work control and work intensity – Weighting structure: Multiple Correspondence Analysis (MCA) + Clustering – Descriptive analysis – Multilevel modelling

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 52

Holm & Lorenz (2015)

• Method

– The results of the factor and cluster analysis provide us with an average characterization of the frequency of the different forms of work organization for the 2000, 2005 and 2010 waves of the survey.

• Discretionary learning: high level of autonomy in work are

combined with high levels of learning, problem-solving, and task complexity

• Lean: lower levels of employee discretion in setting work methods

combined with high use of job rotation and team work and work efforts constrained by quantitative production norms and by the collective nature of work organization

• Taylor: low discretion and low level of learning and problem-solving

combined with the use of teams and job rotation at about average levels

• Simple: traditional forms of work organization where methods are

for the most part informal and non-codified

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 53

Holm & Lorenz (2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 54

Holm & Lorenz (2015)

A striking result: the decline in the frequency of the DL forms

• rganisation over a period of time in which a major objective of

EU was to increase the capacity of firms for learning, creativity, and innovation with a view to making Europe the most dynamic knowledge-based economy in the world

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 55

Holm & Lorenz (2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 56

Holm & Lorenz (2015)

• Method

– Multilevel modelling with three level covariates:

• Employee-level covariates

– Gender, tenure at current workplace, whether or not the worker has undergone vocational training in the prior year, occupation, sector and size of the respondent’s main work place

• Country level covariates

– The mean value of the variable over the three waves – This provides a time-invariant measure of differences across nations

• Country-wave level covariates

– The value observed for the country-wave minus country-level mean value – It reflects intra-country deviations from the country’s time-invariant value and can be interpreted as intra-country variations in contextual effects

– Country-wave and country level covariates finally included in the model result from factor analysis on the initial macro-variables – The dependent variable is a binary variable indicating whether the respondent of the interview has his/her work organized according to the model in focus

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 57

Holm & Lorenz (2015)

Title/date

www.inclusivegrowth.be www.inclusivegrowth.be 58

Holm & Lorenz (2015)

• Conclusions
• Descriptive results

– The dominant trend for EU-15 nations (2005-2010) is toward lower levels of learning and problem-solving and a reduction in the discretion (France, Sweden and Ireland) – For the new member nations, the dominant trend over 2005-2010 is toward an increase in the DL forms and a decline in the Lean and Taylorist forms

• Econometric results:

– On level 3 the significant of inclusive training is observed

• Likelihood of an employee being engaged in the DL forms is higher in

nations that combine high-level expenditure on labor market protection policies with an emphasis on further training

– On level 2 this is no more the case: only economic conjunction is significant

• The likelihoods of the different forms of work organization appear to be

insensitive to changes in the policy variables within nations over time

• Downturns lead to a decline in the frequency of use of the DL forms of work
• rganization

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 59

Holman et al. (2015)

• Aims

– Examining trends – Examining convergence and divergence – Explaining change over time

• Time:

– 1995 – 2000-2005

• Data:

– EWCS

• Country coverage:

– EU-15

• Method

– Four dimensions of quality of working life (objectivist approach)

• Skills and discretion
• Work risks
• Work intensity
• Working time quality

– Weighing structure: summarizing corresponding items – Descriptive analysis – Latent growth modelling of pseudo-panel data

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 60

Holman et al. (2015)

• Method

– Pseudo-panel data

• Form cohorts around a characteristic (country, occupation)
• Treat average score of cohort as observation i.e. panel data
• f cohorts
• 15 EU countries, 9 occupational groups = 135 cohorts

– Latent growth modelling (LGM)

• Enables the analysis of growth over time, with a particular

advantage being the ability to test whether growth is systematic or due to chance

• Each observed case has a specific starting point (the

intercept) and a specific rate of change over time (the slope)

• It can test whether growth varies according to different

starting points and whether growth varies by covariates

Title/dateAcross countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 61

Holman et al. (2015)

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 62

Holman et al. (2015)

• Skill use and development was evaluated using a measure of cognitive

demand, which assessed the extent to which a job requires dealing with complex tasks, and a single item measure on whether training to develop skills has been undertaken

• The job discretion measure assessed the extent to which the person has

control over how to work and the pace of work

• High scores represent high level of job quality

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 63

Holman et al. (2015)

In EU-15, job discretion showed a statistically significant downward trend from 1995 to 2010

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 64

Holman et al. (2015)

In 10 of the EU-15 countries the level of job discretion was significantly lower in 2010 than in 1995

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 65

Holman et al. (2015)

There is no any general pattern of convergence or divergence in job discretion across the EU15,as there was no statistically significant relationship between the initial mean level and rate of change in job discretion

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 66

Holman et al. (2015)

We can see the divergent trends in job discretion according for countries with lower than average trade union density and higher than average trade union density

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 67

Holman et al. (2015)

Social democratic regimes had significantly different rates of growth in job discretion than all other regimes in the EU15, such that the level of job discretion was maintained in social democratic regimes and is diverging from all other regimes in the EU15

Across countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 68

Holman et al. (2015)

• Conclusions

– Trends

• Statistically significant upward trends occurred in training and

working time quality

• Downward trends in job discretion and cognitive demand
• Growth of physical demands and workload

– Convergence and divergence

• No clear pattern convergence or divergence across all

aspects of job quality

• Job discretion diverged between countries partly as a result
• f slower declines in countries with high union density
• The institutional regimes of Nordic countries appear to

protect the quality of job discretion and cognitive demand and a decline in these job quality dimension was evident of other country groupings.

Title/dateAcross countries & over time/10.05.2016

www.inclusivegrowth.be www.inclusivegrowth.be 69

General conclusions

• Objectivist approach is globally preferred to

subjectivist one

• Using several quality of work indicators is globally

preferred to one unique indicator

• Particular interest in context and institutional variables

that are considered to shape quality of working life across countries with however mixed empirical results

• The issue of number of countries available should be

better taken into account when interpreting the results using the multilevel modelling

• Decline in job discretion (work complexity, DL work
• rganisation) in EU-15 between 1995 – 2010 was

found in three of four empirical studies

Across countries & over time/10.05.2016

TÁRKI Social Research Institute Inc. (HU) Amsterdam Institute for Advanced labour Studies, Universiteit van Amsterdam (NL) The Swedish Institute for Social Research, Stockholms Universitet (SE) Fachbereich IV, Wirtschafts- und Sozialstatistik, Universität Trier (DE) Centre d’Etudis Demogràfics, Campus de la Universitat Autònoma de Barcelona (ES) Centre d’Etudes de Population, de Pauvreté et de Politiques Socio-Economiques (LU) Centre for Social Policy, Universiteit Antwerpen (BE) Institute for Social & Economic Research, University of Essex (UK) Bremen International Graduate School of Social Sciences, Universität Bremen (DE) Department of Dynamics of Organisations of Work, Centre d’Etudes de l’Emploi (FR) The Centre for European Policy Studies (BE) Dipartimento di Economica e Menagement, Università di Pisa (IT) Social Statistics Division, University of Southampton (UK) Luxembourg Income Study, asbl (LU) WageIndicator Foundation (NL) School of Social Sciences, The University of Manchester (UK)

Partners Co-ordinator

Inclusive Growth Research Infrastructure Diffusion Contract No 312691 For further information about the InGRID project, please contact inclusive.growth@kuleuven.be www.inclusivegrowth.be p/a HIVA – Research Institute for Work and Society Parkstraat 47 box 5300 3000 Leuven Belgium

Guy Van Gyes Monique Ramioul

InGRID