AboutSubmit my research

Dynamic microsimulation

Research articles · Dec 31, 2020

IrpetDin. A Dynamic Microsimulation Model for Italy and the Region of Tuscany

Maria Luisa Maitino, Letizia Ravagli, Nicola Sciclone

Istituto Regionale Programmazione Economica della Toscana, Florence, Italy

Volume 13(3) Winter 2020

Cite as: M. Maitino, L. Ravagli, N. Sciclone; 2020; IrpetDin. A Dynamic Microsimulation Model for Italy and the Region of Tuscany; International Journal of Microsimulation; 13(3); 27–53. doi: 10.34196/ijm.00224

Abstract

IrpetDin is a dynamic microsimulation model, developed by IRPET (Regional Institute for Economic Planning of Tuscany) to study the future socio-demographic structure of the population and to evaluate the effects of social security programmes in Italy and in Tuscany over the medium to long-term. The model, based on the Eurostat Survey on Income and Living Conditions, makes projections from 2009 to 2050 and it is organised in modules: demography, education, labour and income and social security. IrpetDin produces realistic projections even for the Region of Tuscany and models education and labour with details. Probabilities and rates are estimated differently for Italy and Tuscany, trough regional administrative data. Education careers are completely simulated, from the choice of secondary school to drop-out, from university enrolment to graduation. Labour supply is endogenously determined while labour demand is driven from IRPET’s macro model. The matching of labour supply and demand is modelled by sector of activity and education, in order to estimate the quantitative and the qualitative mismatch.

This article contains 13 figures and 10 tables.

Figures

Figure 1. Italian population by age. Italy.
Figure 2. IrpetDin’s structure.
Figure 3. IrpetDin’s alignements to Dante.
Figure 4. Validation of demography. Year 2017. Italy.
Figure 5. Validation of labour market indicators. Italy.
Figure 6. Demographic dependence for Italy and Tuscany (%).
Figure 7. Education. Italy.
Figure 8. Labour supply and unemployment. Italy.
Figure 9. Quantitative and qualitative mismatch. Italy.
Figure 10. Pensions flows and stocks. Italy.
Figure 11. Pensions: intra-generational equity. Italy.
Figure 12. Pensions: financial sustainability and inter-generational equity. Italy.
Figure 13. Social assistance for old people and in-kind services.

Tables

Table 1. Validation of pensions. Italy
IrpetDin ISTAT Ratio IreptDin ISTAT
Stock of retirees in 2017 10,800,000 11,039,137 0.98
Pension flows 2016-2017-2018 284,138 291,115 0.98
Pension expenditure (billion euro) in 2017 236 232 1.01

Source: IrpetDin, ISTAT.

Table 2. Exogenous indicators/variables
Indicator/variabile External source
Demographic trends Life expectancy ISTAT forecast, central scenario
Fertility ISTAT forecast, central scenario
Migration flows ISTAT forecast, central scenario
Economic trends GDP real growth rate Dante until 2035, then our assumption
GDP nominal growth rate Dante until 2035, then our assumption
Standard Labour Units Dante until 2035, then our assumption
Ratio SLU/employed Our assumption
Labour income growth rate Our assumption
Part time work Our assumption
Social security programmes Pensions and social assistance for old-people As current legislation
Health and long term care As current legislation
Health and LTC costs growth rate Our assumption
Table 3. Unemployment rate, percentiles of replacement rates for employees and employed, Gini among retirees
2018-2032 2033-2042 2043-2050
Basic scenario Unemployment 11,3% 4,6% 0,1%
p25 employees 62,4% 57,8% 54,0%
p75 employees 78,5% 74,2% 72,4%
p25 self-employed 42,5% 34,0% 30,1%
p75 self-employed 53,4% 43,7% 42,2%
Gini 0,322 0,305 0,305
Official scenario Unemployment 9,7% 6,8% 5,4%
p25 employees 62,4% 58,0% 54,2%
p75 employees 78,5% 74,3% 72,5%
p25 self-employed 42,5% 34,1% 30,2%
p75 self-employed 53,4% 43,8% 42,2%
Gini 0,322 0,305 0,305
Table A1. Main features of IrpetDin’s modules
EVENT POTENTIAL CANDIDATES PROBABILITIES ESTIMATION METHOD VARIABLES USED TO DETERMINE EVENTS DATA SOURCE FOR PROBABILITIES ESTIMATION DATA SOUCE FOREVENTUAL ALIGNEMNTS
Ageing All individuals  
Mortality All individuals Rates taken from external sources Territory, age, gender Mortality tables ISTAT (2008-16)  ISTAT forecast, central scenario (2008–2050)
Marriage Single, divorced, widowed aged 18–59 Our calculated rates Territory, age, gender, education Official data on marriage ISTAT (2008–2013)  
Fertility Married/cohabitant women aged 18–45 Our calculated rates Territory, age, n° children, education, nationality Birth attendance certificates RT (2007–2013) + ISTAT survey on births (2012)  ISTAT forecast, central scenario (2008–2050)
Dissolution Married/cohabitant aged 20–64, at least 3 years of marriage Our calculated rates Territory, age, gender, nationality Official data on civil status ISTAT (2008–2013)  
Leaving home Individuals aged 18–59, unmarried, employed, not the head of the family Our calculated rates Territory, age, gender Survey “Famiglia e soggetti sociali” ISTAT (2013)  
DEMOGRAPHY Migration flows All individuals Our calculated rates Territory, gender, education, occupational status, type and size of the family Individual data on registrations and cancellations ISTAT + Demographic balances of foreign citizens ISTAT (2009–2017) ISTAT forecast, central scenario (2008–2050)
Choice of secondary school Individuals aged below 16 Our estimation with multinomial logit Territory, gender, parents’ education Survey on secondary school graduates ISTAT (2011; 2015)  
Educational attainments at secondary school (drop-out, repeating, high school certificate) Enrolled to 1° year of secondary school Our calculated rates and estimation with multinomial logit Territory, gender, parents’ education, type of secondary school School register RT (2008-13) + Survey on secondary school graduates ISTAT (2011; 2015)  
Entry to tertiary school Individuals with secondary school diploma Our calculated rates Territory, gender, type of secondary school, mark, year of study University register (2008–2013) + Survey on secondary school graduates ISTAT (2011; 2015)  
EDUCATION University career (drop outs, three- and five-year degree) Enrolled to university Our calculated rates Territory, age, gender, type of course Survey on university graduates ISTAT (2011; 2015)  
Entry in the labour force Individuals leaving the school (aged 15–39) and inactive people Calculated Territory, gender, age, education, role within HH Labour Force Survey ISTAT (2009–2013; 2014–2016)  
Employment status Individuals belong to the labour force Matching between labour demand (Dante) and labour supply (IrpetDin) Territory, education e sector INPS, data on hours of redundancy funds and Unioncamere – Minister of Labour, Excelsior survey. (2008–2014) Labour demand aligned to Dante 2008–2035, then our assumption
Career employment All individuals employed Our calculated rates Sector Labour Force Survey ISTAT (2009–2013)  
LABOUR AND INCOME Wages and earnings All individuals employed Our estimation with OLS Territory, age, gender, contributory seniority, educational level, work status, number of hours worked, citizenship EU-SILC ISTAT (2003–2013)  
SOCIAL SECURITY Retirement All non-pensioners accruing retirement requirements   Pensions rules    
Pension amount All pensioners   Pensions rules    
Social pension Individual aged above 65 with economic condition requirements   Pensions rules    
Integration at minimum pensions and pension supplements Pensioners fulfilling age and economic condition requirements   Pensions rules    
Health All individuals (insurance value approach) Age, gender, education, nationality Regional administrative data on specialist, pharmaceutical and hospital services (only Region of Tuscany) (2011)  
Long Term Care All individuals Our estimation with logit Age, gender and education Survey “Multiscopo”, ISTAT (2014)  
Table A2. Multinomial logit of high school choice
Italy Tuscany
lyceum (base)   Coef. P>z Coef. P>z
technical          
  female –1,277 0,00 –1,408 0,00
  father with secondary education –0,503 0,00 –0,264 0,00
  father with tertiray education –1,398 0,00 –1,704 0,00
  mather with secondary education –0,675 0,00 –0,742 0,00
  mather with tertiray education –1,613 0,00 –1,531 0,00
  intercept 1,051 0,00 0,998 0,00
professionalising          
  female –0,795 0,00 –1,018 0,00
  father with secondary education –0,850 0,00 –0,817 0,00
  father with tertiray education –1,849 0,00 –1,412 0,00
  mather with secondary education –0,946 0,00 –1,130 0,00
  mather with tertiray education –1,916 0,00 –2,163 0,00
  intercept 0,308 0,00 0,479 0,00
others          
  female 0,133 0,00 0,238 0,00
  father with secondary education –0,519 0,00 –0,512 0,00
  father with tertiray education –0,896 0,00 –1,566 0,00
  mather with secondary education –0,288 0,00 –0,333 0,00
  mather with tertiray education –0,544 0,00 –0,379 0,00
  intercept –2,189 0,00 –2,049 0,00

Source: our estimation on survey on secondary school graduates ISTAT (2011; 2015).

Table A3. Multinomial logit of high school mark
Italy Tuscany
under 70 (base)   Coef, P>z Coef, P>z
70–80  
female 0,383 0,00 0,371 0,00
father with secondary education 0,159 0,00 0,178 0,00
father with tertiray education 0,316 0,00 –0,106 0,04
mather with secondary education 0,098 0,00 0,174 0,00
mather with tertiray education 0,266 0,00 0,283 0,00
intercept –0,739 0,00 –0,536 0,00
80–90  
female 0,639 0,00 0,792 0,00
father with secondary education 0,191 0,00 0,501 0,00
father with tertiray education 0,499 0,00 0,469 0,00
mather with secondary education 0,230 0,00 0,006 0,88
mather with tertiray education 0,483 0,00 –0,158 0,02
intercept –1,518 0,00 –1,404 0,00
90–100  
female 0,789 0,00 0,584 0,00
father with secondary education 0,501 0,00 0,258 0,00
father with tertiray education 0,974 0,00 –0,030 0,65
mather with secondary education 0,482 0,00 0,809 0,00
mather with tertiray education 0,827 0,00 1,229 0,00
intercept –2,168 0,00 –2,113 0,00

Source: our estimation on survey on secondary school graduates ISTAT (2011; 2015).

Table A4. Logit of the being a part-timer
2009-2013 2018-2019
Coef. Std. Err. Wald Chi-Square Pr > ChiQuadr Coef. Std. Err. Wald Chi-Square Pr > ChiQuadr
Woman 2,0279 0,0076 71247,14 <.0001 1,781 0,0106 28357,22 <.0001
Age –0,017 0,000311 2995,31 <.0001 –0,0165 0,000423 1517,71 <.0001
Number of children 0,216 0,00484 1991,45 <.0001 0,1288 0,00648 394,59 <.0001
Intercept –2,2154 0,0148 22545,18 <.0001 –1,7406 0,0209 6913,29 <.0001
Number of obs     791,126   307,954  

Source: our estimation on Italian Labour Force survey 2009-2013 2018-2019, ISTAT.

Table A5. OLS of the logarithm of income from employee work
Coef. Std. Err. t Pr > |t|
Intercept 7.70048 0.01596 482.35 <.0001
Age 0.04376 0.000727 60.15 <.0001
Age(squared) –0.0004 8.74E-06 –46.05 <.0001
Man 0.18824 0.00259 72.54 <.0001
Primary education –0.12254 0.00469 –26.14 <.0001
Tertiary education –0.16541 0.00875 –18.91 <.0001
Tertiary Education (squared) 0.00664 0.000213 31.15 <.0001
Exceutive 0.52652 0.00418 125.91 <.0001
Office worker 0.28028 0.00268 104.45 <.0001
Head of the household 0.07564 0.00248 30.46 <.0001
Women with children<18 –0.00165 0.00482 –0.34 0.7327
Ptime 0.14081 0.0055 25.62 <.0001
Working hours 0.0194 0.000148 130.86 <.0001
R2 0.4226

Source: our estimation on EU-SILC2008. ISTAT.

Table A6. OLS of the logarithm of income from self-employed work
Coef. Std. Err. t Pr > |t|
Intercept 8.52019 0.03519 242.15 <.0001
Age 0.01755 0.00116 15.07 <.0001
Age(squared) –5.5E-05 1.18E-05 –4.67 <.0001
Secondary Education 0.1194 0.01231 9.7 <.0001
Tertiary education –0.0409 0.03009 –1.36 0.1741
Tertiary education (squared) 0.00904 0.000563 16.06 <.0001
Professionals 0.01318 0.01036 1.27 0.2036
Head of the household 0.14279 0.0074 19.29 <.0001
Services –0.02719 0.00781 –3.48 0.0005
Man 0.20699 0.00779 26.59 <.0001
Working hours 0.00615 0.00029 21.21 <.0001
R2 0.1549

Source: our estimation on EU-SILC2008. ISTAT.

Table A7. Logit of the probability of being not self-sufficient
Coef. Std. Err. z P>z [95% Conf.Interval]
Tertiary education –1.110 0.635 –1.750 0.080 –2.354 0.134
Woman 0.652 0.149 4.360 0.000 0.359 0.945
Age 0.091 0.008 11.650 0.000 0.076 0.106
Constant –8.845 0.579 –15.290 0.000 –9.979 –7.711
Number of obs 7.049Pseudo R2 0.288

Source: our estimation on Survey “Multiscopo”, ISTAT (2014).