1. OBJECTIVES, INSTRUMENTS AND DATASET
Marco Cacciotti and Desislava Dankova
IESS – Intermediate Workshop
Rome - 27th of November, 2015

2. Outline Project’s objectives: intermediate and final
Instruments: Models, data and software
Focus on the Dataset: update and extension

3. Overall purpose of the project IESS
objectives
Overall purpose of the project IESS
The project «Improving effectiveness in social security» (IESS) aims at assessing the cost effectiveness of the Italian Pension System in the short and long run focusing on its sustainability, adequacy and policy analysis.
IESS follows up a previous project named «Innovative Datasets and Models for Improving Welfare Policies».
It will provide innovative analytical tools to improve the effectiveness of public policy evaluation.

4. Main aspects of IESS Intermediate objectives Fields of interest Focus
Tools/Results
1. Sustainability
Effects on employment and productivity of ageing and increasing retirement ages
Literature Survey
Intermediate objectives
2. Adequacy
A. Analyses of workers’ career paths
B. Public Pension benefits for workers with discontinued careers
C. Italian private second and third pension pillars
Micro-econometric analyses.
3. Policy analysis
A. Jobs acts and pension system.
B. Workability and Labour supply of pensioners.
C. Social determinants of health status and life expectancy
Micro-econometric analyses.
Final objective

5. Innovations vis-à-vis the previous project
objectives
Innovations vis-à-vis the previous project
Extension and improvement of the dynamic microsimulation model T-DYMM;
Updating the existing modules;
Endogenous parameters for long run projections;
Inclusion of complementary pension savings;
Improvement of the estimates relative to the labour market;
Updating and extension of the dataset used for the microsimulation by developing a new version of AD-SILC (Administrative SILC)

6. Main tools used in the microsimulations
instruments
Main tools used in the microsimulations
A new microsimulation platform LIAM2 ensuring increasing benefits in terms of speed and data capacity.
A new version of the dataset AD-SILC allowing to carry out more robust analyses about the individual dynamics on the labour market and the pension accumulation in the contributory scheme and to supply a more detailed database to the micro-simulation model TDYMM.

7. Administrative vs survey datasets
Administrative data (INPS)
Advantages: very wide coverage and very long time span
Limits: only information needed for administrative modules
Survey data (SILC)
Advantages: rich information about individuals’ background and socio-economic characteristics
Limits: very limited sample size and short time span

8. Merging administrative with survey data
dataset
Merging administrative with survey data
AD-SILC
(administrative SILC)
Information about all workers paying contributions to INPS funds, gross wages, unemployment benefits, sickness allowances and maternity allowances; the workers’ qualification, typologies of work
Large number of socio-economic variables at household and individual level
Very large time span

9. Structure of the old version of AD-SILC
dataset
Structure of the old version of AD-SILC
IT-SILC 2005
56,105 individuals
AD-SILC
1,162,045 observations
56,105 individuals
43,388 registered at least once in the INPS archives
INPS: working history of individuals surveyed in SILC and registered in INPS archives
Longitudinal data in the period

10. Structure of the new version of AD-SILC
dataset
Structure of the new version of AD-SILC
AD-SILC
3,355,801 observations
202,725 individuals
61,542
SILC 04 A4 B3 C2 D1
56,105
SILC 05 B4 C3 D2 E1
54,512
SILC 06 C4 D3 E2 F1
52,772
SILC 07 D4 E3 F2 G1
52,433
SILC 08 E4 F3 G2 H1
51,196
SILC 09 F4 G3 H2 I1
47,551
SILC 10 G4 H3 I2 J1
47,841
SILC 11 H4 I3 J2 K1
47,365
SILC 12 I4 J3 K2 L1
INPS: working history of individuals surveyed in SILC waves and registered in INPS archives
Longitudinal data in the period
Number of observations: 5,016,214 number of individuals: 147,777

11. dataset
Contents and features
AD-SILC is an unbalanced 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.

12. Advantages of the new AD-SILC
dataset
Advantages of the new AD-SILC
Greatly enlarged sample size to analyze more in details smaller and specific subgroups of Italian labour force.
Extending the dataset up to 2014 allows to study in comprehensive way the weaknesses engendered by the current macroeconomic crisis and its impact on labour market dynamics and income distribution.
Updating the base data as a starting point for the simulations.

13. Regressions and projections
dataset
Regressions and projections
Regressions used in the model are based on the entire dataset AD-SILC.
All individuals in IT-SILC and the respective working and contribution history carried out by INPS are considered on the period
The reference year of the simulations is 2011, where the data used are a single extract of AD-SILC.
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.

14. Annual composition by gender and sample share by year
dataset – descriptive statistics
Annual composition by gender and sample share by year
Source: elaborations on AD-SILC data
The large increase in the number of observations since 2005 is due to the fact that information provided by the Register of Retired is available only since that date.

15. Trend of employment status (absolute values)
dataset – descriptive statistics
Trend of employment status (absolute values)
Source: elaborations on AD-SILC data

16. Employment composition (percentage values)
dataset – descriptive statistics
Employment composition (percentage values)
Source: elaborations on AD-SILC data

17.
Il progetto IESS è finanziato dal Programma per l’Occupazione e la Solidarietà Sociale dell’Unione Europea – PROGRESS ( ).
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.