Max Planck Institute for Human Development Choices, Constraints or Preferences? Identifying Answers from Part-time Workers Transitions CCSR Seminar Series.

Max Planck Institutefor Human Development

Choices, Constraints or Preferences? Identifying Answers from Part-time Workers’

Transitions

CCSR Seminar Series

14th of November, 2006

VANESSA GASH


What is the issue?

Why is that women work part-time and men don’t?

1 - Women prefer part-time jobs - preference

or

2 – Women cannot work full-time given household obligations – constraint.

(Hakim 1991, 2000, 2002; McRae 1991, 2003; Fagan and Rubery 1996, Fagan 2001, Warren 2001, Ginn et al 1996…etc etc.. ).


Aims of the Research Paper

Empirical analysis of women part-timers’ labour market transitions.Do part-timers reveal different LM trajectories and if so why?

Three country analysis. Does institutional context structure the transitions of part-time workers?

Previous research has failed to compare countries which have a strong public childcare component (Drobnic, Blossfeld and Rohwer 1999; O’Reilly and Bothfeld 2002).

Women’s access to childcare is considered pivotal to their ability to work full-time


Review of the preferences versus constraint debate

Review of the institutional factors likely to structure part-time workers transitions

Results. Non-parametric and parametric analysis of female part-time workers’ transitions to; full-time employment, to inactivity and to unemployment.

Conclusions

Presentation Outline


Individual action is a function of individual choice.

Preference Theory: Preferences determine social outcome (Hakim, 2002).

Attribute unexplained differences to preferences (Barrett and Doiron 2001; Petrongolo 2004).

Questionable to attribute observed differences to unmeasured preferences.

Part-time as chosen.


Operationalised by three questions:

1- Preferred family employment structure?

2- Who is the main income earner in your household?

3- If you had enough money to not work, would you still work?

WORK-CENTERED, ADAPTIVE, HOME-CENTERED.

Hakim’s Preference Theory


Part-time means of reconciling paid work and family responsibilities.

Preferences product of experience.

Psychological process of cognitive dissonance (Festinger 1957).

Can preferences be met?

Part-time as Constraint


1- Child and elder care. State investment? DK- 1.96 of GDP, Fr- 1.5 of GDP, UK – 0.31GDP

2- Full-time working hours compatible with social/home life.

Proportion of workers working more than 45 hours per week (OECD-2001).DK: 25% of men, 7% of women.FR: 18% of men, 7% of women.UK: 43% of men, 12% of women.

3- General institutional compatibility of full-time (p/m)aternal employment, i.e. school hours, shop-opening times.

Which institutions might allow women to be true

chosen part-time workers?


Comparative Indicators: Female Employment

Female Employment Rate

(OECD)

Proportion Part-time(OECD)

Quality of Part-time jobs

1994 - 2001 1994 - 2001

% % Wage penalty

Low Skilled Jobs

Denmark 67.1 - 71.4 25.6 - 21.0 no yes

France 51.5 - 55.2 29.9 - 24.4 no no

United-Kingdom 62.1 - 66.1 40.7 - 40.1 yes yes


The Analysis

Part-timers transitions in three countries which provide different opportunity structures for part-time work to be a chosen.

1- Denmark and France provide workers with access to childcare.

2- Denmark and France provide better quality part-time work.

3- Gender contract more egalitarian in Denmark and France, institutional structures more supportive of full-time.


The Data

European Community Household Panel (1994-2001)

The ECHP is a standardised comparative cross-national survey.

The samples were drawn by each member state as simple random samples, with information collected from respondents in face-to-face interviews in each panel year (1994-2001).

The data set contains information both at the individual and household levels relating to human capital acquisition, occupation and industrial location.


The Data

Wave 1 (1994)

Household Interviews

Wave 1 (1994)

Personal Interviews

Response rate W1

Attrition rate

W1 and 2

Denmark 3,482 5,903 64.2% 11%

France 7,344 14,331 79.7% 10%

United-Kingdom

(not BHPS)

5,779 10,517 71.3% 23%

EU (12) 60,819 128,043 73%


Modelling Strategy

• The statistical technique applied, event history analysis, allows one examine the transition rates of part-time workers.

• Key statistical concept is of the hazard/transition rate: “conditional likelihood that an event takes place during a time interval t -> t+1, conditional on it not having occurred before t”.

• Dependent variable is the duration of the individual in a part-time job.

• Dep. variable was constructed using information on job-start and job-end dates. NO calendar data in ECHP that distinguishes between types of employment.

• Allows for a distinction between within and between job changes.


Part-time Workers Transitions

Three Transitions:

1- Part-time (<30 hours) to Full-time (+30 hours).

2- Part-time to Inactivity (Housewive/other economically inactive).

3- Part-time to Unemployment.


Measuring Preferences and Constraints

Constraint:

Time varying categorical variable of the number of children within the household. (1= 0-3, 4-12, 13-18 years).

EXPECT: N of Children to constrain PTs trsn to FT in countries with little/no childcare (UK).

Preferences:

Difficult to measure but “reasons” for PT used as a proxy. (Choice, Housework, Under-employed, other)

EXPECT : Chosen part-time to make fewer transitions to full-time (Expected to be true of all countries).


Transition from PT to Full-time: Hypotheses

H1: Different outcomes by preferences, with chosen part-timers LESS likely to make transitions to full-time.

H2: UK part-time workers’ transitions more likely to be constrained, with constraints (as a result of the institutional structure) over-riding preferences.

H3: DK part-time transitions are most likely to be a function of preference, with institutional structure offering worker-carers a real choice.

H4: Constraining effect of children within the home on transitions to full-time in the UK only (where institutions are expected to constrain workers).


Transition from PT to INACTIVITY: Hypotheses

H1: Preferences are likely to be predictive of transitions to inactivity.

H2: Number of children within the home expected to increase transition rate in countries with little public childcare (UK).


Transition from PT to Unemployment: Hypotheses

H1: Preferences should have NO impact on transitions to unemployment.


Difference in the Kaplan-Meier Survival Estimates of Part-time Workers' Transtions versus Full-time Workers' Transtions to FULL-TIME

EMPLO YMENT

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

Time in Months (0-96)

Dif

fere

nce

in P

ropo

rtio

n Su

rviv

ing

a T

rans

ition

to F

ull-

time

DK FR UK


Difference in the Kaplan-Meier Survival Estimates of Part-time Workers' Transtions versus Full-time Workers' Transtions to INACTIVITY

-0.3

-0.2

-0.1

0

0.1

0.2

0.3


Dif

fere

nce

in

Pro

po

rtio

n S

urv

ivin

g a

Tra

nsi

tion

to I

NA

CT

IVIT

Y

DK FR UK


Difference in the Kaplan-Meier Survival Estimates of Part-time Workers' Transtions versus Full-time Workers' Transtions to

UNEMPLO YMENT

-0.3

-0.2

-0.1

0

0.1

0.2

0.3


Dif

fere

nce

in P

ropo

rtio

n Su

rviv

ing

a T

rans

itio

n to

Une

mpl

oym

ent

DK FR UK


Model Strategy

- Piecewise constant exponential model of the transitions of female part-time workers to full-time, inactivity and unemployment.

- Controlling for: educational level (3 cat), skills training, age (5 cat ), unemployment experience, occupation (5 cat), firm size (4 cat) and industrial sector.


Piecewise Constant Exponential Model: PT to FT

Denmark

France United-Kingdom

Reasons for PART-TIME: CHILDCARE

- ns ns

Reasons for PART-TIME: CHOSEN

ns ns -

Reasons for PART-TIME: OTHER

- ns ns

(REF: UNDER-EMPLOYED)

Children aged (1-3yrs)

+ ns -


ns ns -


+ ns ns

(Ref – no children


Piecewise Constant Exponential Model: PT to INA

Denmark



ns ns ns


ns + ns


ns + ns



ns + ns


ns ns ns


ns - ns



Piecewise Exponential Model: PT to UNEMP

Denmark



ns ns ns


ns ns ns


ns ns ns



ns ns ns


ns ns na


ns ns +



Conclusion

PT-INA

All part-time more time dependant with UK part-time the most likely to make these transitions.

PT-UNEMP

DK and FR part-time employment disproportionately exposed to

unemployment.

Strong cross-national differences in part-time workers outcomes.

PT-FT

UK part-time employment is time dependant suggesting: either strong preferences or strong constraints.

No evidence of either for DK or FR, where the S(F) were the same.


Conclusion

Preferences/Reasons for part-time are of limited value in predicting future transitions.

They do „work“ in the right direction, in some countries, but not sufficiently to suggest that they explain why women work part-time.

Of most value in explaining transitions to full-time, of no value to predict transitions to unemployment.


Conclusion

Institutional context structures outcomes.

N of children had NO negative impact on part-timers transtions to full-time in countries with public childcare and/or little long hours working culture, DK and FR.

In the UK, children constrained PTs trsn to FT.

N of children did however increase French part-timers transtions to inactivity, but only for very young children.

N of children found to increase UK women‘s unemployment risk.


Thanks for listening!

Max Planck Institute for Human Development Choices, Constraints or Preferences? Identifying Answers from Part-time Workers Transitions CCSR Seminar Series.

Documents

max planck institute

human development review

fulltime slide

human development aims

human development choices

constraint slide

true chosen parttime

fulltime employment