The Labour Market Module:

data, estimations and results

Desislava Dankova

IESS – Final Conference

Rome - 16th of May, 2016

2

AD-SILC dataset: contents and features

• AD-SILC is an unbalanced panel dataset containing both:

• retrospective information on individuals’ working conditions

before the year of survey of SILC, and

• forward-looking information on individuals’ working

conditions after the year of survey of SILC.

• Panel INPS - longitudinal data of individuals’ working history

since their entry in the LM: occupational status, income

evolution, contribution accumulation, etc.

• Panel SILC - longitudinal data of individual socio-economic

characteristics (up to 4 years): education, marital status, number

of children, etc.

3

Analyses, regressions and projections (1)

• Analyses of the workers’ dynamics in Italy – evidence from

the AD-SILC dataset:

• Transition matrixes

• Earnings distribution trends

• Accumulation of pension contributions

4

Analyses, regressions and projections (2)

• Regressions used in the model are based on the entire

dataset AD-SILC.

All individuals in IT-SILC 2004-2012 and the respective

working and contribution history carried out by INPS are

considered over the period 1998-2011.

• Modelling the demographic dynamics

• Modelling the working statuses

• Modelling the earnings process

5

Analyses, regressions and projections (3)

• Simulations – based on a single extract of AD-SILC.

• 2011 is the starting point of the simulation, with a sample

which is representative of the Italian population in that year.

• The dataset is cross-sectional, integrated with retrospective

information about working conditions, acquired work

experience, total number of years of contribution, etc.

The base sample of the model includes individuals surveyed

in SILC 2011 and the respective working, labour income and

contribution conditions registered in INPS archives.

6

Some evidence on the labour market

conditions in Italy from the AD-SILC

dataset

7

LM transitions

Perm Fixed Self-empl. Atypical Out of work

Transitions between working statuses after 1 year – 2000 vs 2008

Open-ended employees

87.8 87.5

94.4 93.7 94.4 95.0

3.4 3.6

1.3 1.7 0.6 1.1 7.3 7.7

3.5 3.8 4.5 3.5

60%

65%

70%

75%

80%

85%

90%

95%

100%

2000-01 2008-09 2000-01 2008-09 2000-01 2008-09

15-34 35-44 45-54

34.0 25.0 22.2 19.3

11.9 16.0

48.1

48.5

64.7

59.0 73.6 65.3

15.4 22.4

11.4 18.1

12.7 16.2

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2000-01 2008-09 2000-01 2008-09 2000-01 2008-09

15-34 35-44 45-54

Fixed-term employees

8

Perm Fixed Self-empl. Atypical Out of work

Transitions between working statuses after 1 year – 2000 vs 2008

Self-employed

2.1 0.9

0.7 1.9 0.3 1.1 0.3 0.7

93.0 90.7 94.9 94.4 95.7 95.5

3.4 5.0 2.2 3.3 3.0 2.9

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2000-01 2008-09 2000-01 2008-09 2000-01 2008-09

15-34 35-44 45-54

10.7 9.1 7.6 4.6 2.5 4.2

5.8 8.8 1.5 5.5 0.0 1.6

62.2 60.6 76.3 79.2 87.9

83.1

18.3 18.7 11.1 8.3 8.8 8.8

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2000-01 2008-09 2000-01 2008-09 2000-01 2008-09

15-34 35-44 45-54

Atypical workers

9

Persistence in the work state in 2008 after 1 and 3 years (by age class)

35-44 45-54 15-34

87.5

48.5

90.7

60.6

81.3

32.8

81.5

37.1

00

10

20

30

40

50

60

70

80

90

100

2008-09 2008-11

93.7

59.0

94.4

79.2

89.5

45.0

89.4

64.6

00

10

20

30

40

50

60

70

80

90

100

2008-09 2008-11

95.0

65.3

95.5

83.1

89.8

53.2

91.2

73.2

00

10

20

30

40

50

60

70

80

90

100

2008-09 2008-11

10

2008 Perm. Fixed term Self-empl. Atypical Out of work

Perm. 95.8 1.4 0.5 0.5 1.8 Fixed 23.0 53.6 2.6 3.8 17.0

Self-empl. 1.0 1.0 96.1 0.4 1.5 Atypical 3.9 8.2 1.7 80.3 6.0

2008 Perm. Fixed term Self-empl. Atypical Out of work Perm. 94.6 1.4 0.5 0.2 3.3

Fixed 19.5 58.8 2.8 1.5 17.5 Self-empl. 0.8 1.1 94.8 0.3 3.0

Atypical 4.1 3.7 2.5 80.4 9.4

2009

2008 Perm. Fixed Term Self-empl. Atypical Out of work

Perm 91.2 2.4 0.5 0.1 5.8

Fixed Term 18.3 60.6 1.9 0.4 18.8

Self-empl. 0.9 1.1 93.0 0.4 4.6 Atypical 7.4 2.8 4.6 76.9 8.3

Working conditions after 1 year of those employed in 2008 (by education)

At most lower-secondary

Upper-secondary

Tertiary

Note: workers aged 35-44 in 2008 are considered

11

Perm Fixed term Self-empl. Atypical Out of work Perm 93.4 2.0 1.2 1.0 2.5 Fixed 37.1 41.1 5.7 4.4 11.8

Self-empl. 3.1 2.0 92.2 0.6 2.1 Atypical 12.6 6.7 4.5 65.9 10.3

2011 2008 Perm Fixed Term Self-empl. Atypical Out of work

Perm 85.0 4.3 1.7 0.3 8.8

Fixed Term 29.4 49.5 2.8 0.8 17.5

Self-empl. 3.5 2.2 86.9 1.0 6.4 Atypical 20.4 6.1 9.2 54.1 10.2

Perm Fixed term Self-empl. Atypical Out of work Perm 91.0 2.6 1.5 0.5 4.5 Fixed 36.9 40.6 3.5 2.3 16.7

Self-empl. 2.8 1.6 90.2 1.0 4.4 Atypical 10.6 3.8 7.2 67.8 10.6

Working conditions after 3 years of those employed in 2008 (by education)

At most lower-secondary

Upper-secondary

Tertiary

Note: workers aged 35-44 in 2008 are considered

12

Trend of yearly gross earnings by work typology

Labour earnings dynamics

0

5000

10000

15000

20000

25000

30000

35000

40000

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

PR employees PB employees Professionals Gest.Sep. Self-empl.

13

Trend of earnings inequality – Gini index

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

PR employees PB employees Professionals Gest.Sep. Self-empl.

14

Contributions accumulation for the first NDC

cohorts

• Focus on individuals having started to work in 1996-1998.

• Followed for 13 years (around 1/3 of the career).

• Accumulation adequacy assessed with respect to a representative individual working continuously as a full-time employee and earning the median wage (around 25,000 real gross annual Euros).

15

Distribution of contribution weeks (wrt potential weeks)

Actual weeks on potential weeks ratio

16

Relative distribution of contributions accumulation (wrt median employee)

Accrued contr. on median employee’s contr. - ratio

17

Modelling the labour market dynamics in

T-DYMM:

features and simulation results

18

1. Probability to be employed (all individuals who are not students nor

retired are included in the regressions);

2. Probability to be atypical worker among all workers defined in step 1;

3. Probability to be an employee among workers defined in step 1

except atypical workers;

4. Probability to be self-employed (residual category );

LM transitions (1)

• Conditional probabilities of LM transitions across employment states are

estimated based on a sequence of binary behavioural choices with

the following logical order:

19

LM transitions (2)

1. Economic sector (private vs public);

2. Contract duration (temporary vs permanent);

3. Time arrangements (part-time vs full-time).

Among employees the subsequent choices are concerned:

20

LM transitions (3)

• Sample size: 1,105,456 observations, relative to 82,137 individuals

aged 16-69 years old.

• Estimation period: 1998-2011.

• The estimations are carried out separately for men and women.

• Random effect logit models for LM transitions in order to account

for individual unobserved heterogeneity.

• Lagged labour states are also included among the regressors.

NB: we do not include in our regressions any variable that is not

present in the “simulation world” because of the impracticability of projecting its evolution in time.

21

0

5

10

15

20

25

30

35

40

45

employee self-employed atypical worker

Work typology at time t-1

mal

esfe

mal

es

0.0

0.5

1.0

1.5

2.0

2.5

3.0

tertiary degree upper-sec. degree

Education

mal

es

fem

ale

s

0.0

0.5

1.0

1.5

2.0

age work experience duration in empl. (lag)

Age and experience

0.0

0.5

1.0

1.5

2.0

2.5

3.0

partner in work(lag)

married children aged 0-3 children aged 4-11

Other individual characteristics

Prob. to be employed conditional on individual characteristics – odds-ratios

22

Employment composition by work typology 0

.2.4

.6.8

1

2005 2015 2025 2035 2045 2055 2005 2015 2025 2035 2045 2055

Males Females

Employees Self-employed Atypical workers

Year

Source:T-DYMM - own elaborations

23

Estimations of earnings

Yearly individual labour income gross of personal income

taxation is the product of two components:

monthly gross wages months worked

The earnings process is

modelled separately for

the three work typologies

and by gender

Modelled in two steps:

1) The probability of being in work

all year (concerns atypical and

temporary workers)

2) Define the months worked for

those workers who are not

assigned to the «work all year»

status

24

Wage function

• The WF consists of a vector of observed variables (𝑿𝒊𝒕) and unobserved

variables which are represented be a random component that captures

heterogeneity in permanent differences between individuals (𝑢𝑖) and a

stochastic error component (ν𝑖𝑡):

– The permanent error component, 𝑢𝑖, (i.e. intellectual ability, soft skills,

motivation) represents a constant wage deviation for each individual,

where 𝑢~𝑁(0, 𝜎𝑢2).

– The transitory component, ν𝑖𝑡 , (i.e. bonuses, illness, overtime)

follows an AR(1) process plus a white noise error, 𝜀𝑖𝑡:

𝑦𝑖𝑡 = 𝑿𝒊𝒕𝜷 + 𝑢𝑖 + ν𝑖𝑡

ν𝑖𝑡 = 𝜌ν𝑖,𝑡−1 + 𝜀𝑖𝑡 , 𝜀~𝑁(0, 𝜎𝜀2) and 𝜌 < 1

25

Estimations of monthly wages

• A random effect GLS estimator has been utilised to estimate the

wage equation on the AD-SILC panel data.

• Estimation period: 1998-2011

• The estimations are carried out separately for the three work

categories and for men and women.

• Sample size: 632,762 observations for 79,009 individuals aged

20-60: about 75% are employees,19,5% are self-employed and

5,5% are atypical workers.

26

Estimations of months worked

1. Estimations of the probability of being in work all year:

• Random Effect Logit model;

• Sample size – 96,933 observations for 29,391 individuals: 48% are

men and 52% are women;

• Estimation period: 1998-2011.

2. Estimations of months worked:

• Same model as for monthly wages;

• Sample size– 50,264 observations for 12,768 individuals: 41% are

men and 59% are women;

• Estimation period: 1998-2011.

27

Monthly wages of employees (estimation results) Males (1) Females (2)

b se b se

tertiary degree 0.545 *** 0.006 0.4411 *** 0.007

upper-sec. degree 0.2088 *** 0.004 0.2027 *** 0.005

age 0.0893 *** 0.003 0.0381 *** 0.005

age2 -0.0022 *** 0 -0.0011 *** 0

age3 0 *** 0 0 *** 0

work experience 0.0227 *** 0.001 0.0241 *** 0.001

work experience2 -0.0003 *** 0 -0.0004 *** 0

years as employee (lag) 0.0082 *** 0 0.0113 *** 0

perm. contract 0.0508 *** 0.003

perm. contract (lag) 0.0137 *** 0.001 0.0371 *** 0.003

part-time -0.3741 *** 0.003 -0.3225 *** 0.003

part-time (lag) -0.0391 *** 0.003 -0.0645 *** 0.003

public 0.1118 *** 0.004 0.1057 *** 0.007

public (lag) 0.0109 *** 0.004 0.0977 *** 0.006

in work (lag) 0.0314 *** 0.002 married 0.0098 *** 0.002 -0.0281 *** 0.004

partner in work 0.0055 *** 0.002 children aged 0-3 -0.1881 *** 0.003

constant 5.9656 *** 0.038 6.482 *** 0.066

σu 0.2812 0.2974 σν 0.1719 0.3242 ρ 0.4638 0.2878

R2-within 0.1955 0.122

R2-between 0.4704 0.4902

R2-overall 0.3998 0.3837

N.obs. 272,072 217,742

28

0

40

00

06

00

00

20

00

0

Avera

ge

an

nu

al gro

ss in

co

me

2005 2015 2025 2035 2045 2055Year

Males Females

Source:T-DYMM - own elaborations

0

200

00

400

00

600

00

Avera

ge

in

co

me

with

GD

P g

row

th

2015 2025 2035 2045 2055Year

Males Females

Source:T-DYMM - own elaborations

Trend of yearly gross incomes by gender

Trend of yearly gross incomes by gender (with GDP growth)

29

Trend of monthly wages of employees (males)

100

02

00

03

00

04

00

0

Ave

rage

mo

nth

ly w

ag

e

2005 2015 2025 2035 2045 2055

15-34 35-54 55-69

Source: T-DYMM 2.0 - own elaborations

Trend of monthly wages of employees (females)

150

02

00

02

50

01

00

03

00

0

Ave

rag

e m

on

thly

wa

ge

2005 2015 2025 2035 2045 2055

15-34 35-54 55-69

Source: T-DYMM 2.0 - own elaborations

30

www.iess-project.eu

Il progetto IESS è finanziato dal Programma per l’Occupazione e la Solidarietà Sociale dell’Unione Europea – PROGRESS (2007-2013).

Le informazioni contenute in questo documento riflettono solamente le posizioni dell’autore. La Commissione Europea non può essere

considerata in alcun modo responsabile dell’uso che può essere fatto di quanto in esso contenuto.