Silence is Golden? Assessing the Public Debate on Pension Reforms in

Europe

CEPS, 14 September 2004

Tito Boeri

Università Bocconi and Fondazione Rodolfo Debenedetti

Outline

• How informed are citizens about the costs of public pensions?

• Press-media coverage of pension reforms• Involvement of citizens in the public debate• Informational content of the public debate• Information and opposition to reforms• Are there better ways to inform?

Key points

• Individuals poorly informed about individual costs and intergenerational redistribution operated by pension systems

• Those informed are more prone to support reforms increasing sustainability

• Press-media coverage not much helpful and may scare people

• We need more “orange envelopes”

How Informed?

• Public opinion surveys in Germany and Italy, 2000, 2001 and 2004 (also France and Spain in 2000).

• Individuals were asked about:

– aggregate costs

– individual costs

– intergenerational redistribution

operated by public pension systems

Aware of the aggregate Budget Constraint? …

According to you, public pension system is...

0

10

20

30

40

50

60

70

80

Italy (2004) Germany (2004) France (2000) Spain (2000)

balanced surplus deficit

Aware of unsustainability?

There is a risk of pension crisis in 10-15 years...

0102030405060708090

100

Italy Germany France Spain

2000 2001 2004

Aware of reforms being parametric?

Past pension reforms.... (2004)

0

10

20

30

40

50

60

70

80

…have stabilized thepension system

…represent just a firststep towardstabilization

…were ineffective

Italy Germany

Aware of individual costs?

Which percentage of your salary is used to pay public pension contributions?

0

10

20

30

40

50

60

underestimate correct overestimate don't know

Germany (2004) Italy (2004) France (2000) Spain (2000)

Aware of intergenerational redistribution (PAYG)?

According to you, for which purpose are the contributions used for? (2004)

0

0.1

0.2

0.3

0.4

0.5

0.6

to pay only yourown futurepension

to pay onlycurrent pensions

to pay both yourown future and

current pensions

refused /don'tknow

Italy Germany

Perceived intergenerational redistribution: a lump of labour….

People in their late 50's should give up work to make way for younger and unemployed

0

50

100

agree

Eurobarometer Survey, 2000

… fallacy!(youth unemployment and early retirement)

Early retirement (exit rate, born 1931-40) and changes in youth unemployment (aged 21-30), Men

dif

fere

nce

in u

ne

mplo

ym

en

t ra

te

exit rate20 40 60 80

-10

0

10

20

Source: Boldrin et al. (1999) using three European Labour Force Surveys for the years 1986, 1991, and 1996.

Press coverage

Trend in Italy

Number of Quotations of “Pension(s)” in «Il Sole 24 Ore»

in titles in articles

v.a. % v.a %

2000, January 91 1.68 407 7.51 2001, January 87 1.61 392 7.24 2002, January 88 1.62 421 7.77 2003, January 110 2.03 414 7.64 2004, January 99 1.83 392 7.24

Note: Number of articles in which “pension or pensions or similar” are quoted in the title or in the full text as a percent of all articles published on January, from 2000 to 2004.

Degree of involvement in the public debate

Which was your level of attention in following the pension reform debate in Italy? (2004)

05

10152025303540

high medium low none

"During the last months, have you read newspaper articles/watched TV programs concerning pension

reform debate?" (Italy 2004)

0

10

20

30

40

50

60

70

yes no

Who decides to be involved?(Italy, 2004)

Probit estimates Number of obs = 1494 LR chi2(8) = 147.92 Prob > chi2 = 0.0000 Log likelihood = -955.58685 Pseudo R2 = 0.0718 ------------------------------------------------------------------------------ attentionl~l | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- male | .2086094 .0676863 3.08 0.002 .0759468 .3412721 age16_25 | -.9919626 .1277953 -7.76 0.000 -1.242437 -.7414884 age26_35 | -.6984737 .0975444 -7.16 0.000 -.8896571 -.5072903 age36_45 | -.3621833 .0902511 -4.01 0.000 -.5390721 -.1852945 age66_80 | -.2758762 .1127961 -2.45 0.014 -.4969525 -.0548 compulsory | -.2329584 .076926 -3.03 0.002 -.3837307 -.0821861 university | .3791137 .1018227 3.72 0.000 .1795448 .5786825 tradeunion | .2431504 .0772632 3.15 0.002 .0917173 .3945835 _cons | .3308777 .0812599 4.07 0.000 .1716113 .4901441 ------------------------------------------------------------------------------

Informational content of the public debate

Correct estimate of contribution rate by level of attention in following pension reform debate

05

1015202530354045

high medium low none

Correct estimate of contribution rate:have you read newspaper articles/ watched TV program concerning pension reform debate?

0

5

10

15

20

25

30

35

40

yes no

Does attention increase information about individual costs?

(Italy, 2004)

Probit estimates Number of obs = 627 LR chi2(8) = 59.69 Prob > chi2 = 0.0000 Log likelihood = -369.93873 Pseudo R2 = 0.0747 ------------------------------------------------------------------------------ correct | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- male | .3385721 .1099259 3.08 0.002 .1231212 .554023 age26_35 | .4054769 .126531 3.20 0.001 .1574808 .6534731 compulsory | -.2090848 .1468436 -1.42 0.154 -.4968929 .0787233 university | .1820134 .1414748 1.29 0.198 -.0952721 .459299 highincome | .2383185 .1197091 1.99 0.047 .003693 .4729441 manager | .4363686 .1940476 2.25 0.025 .0560423 .816695 clerk | .4091622 .1407087 2.91 0.004 .1333781 .6849462 attention | .3142244 .1317166 2.39 0.017 .0560646 .5723841 _cons | -1.329374 .1854585 -7.17 0.000 -1.692866 -.9658821 ------------------------------------------------------------------------------

Does attention increase information about intergenerational redistribution (PAYG)

(Italy, 2001 2004)

Probit estimates Number of obs = 2779 LR chi2(5) = 87.33 Prob > chi2 = 0.0000 Log likelihood = -1866.4727 Pseudo R2 = 0.0229 ------------------------------------------------------------------------------ infpayg | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- male | .1282726 .0493216 2.60 0.009 .0316041 .2249411 age16_35 | -.175351 .0517041 -3.39 0.001 -.2766892 -.0740128 employed | .3598553 .0579328 6.21 0.000 .2463092 .4734015 a2004 | .0820258 .0612149 1.34 0.180 -.0379532 .2020049 attention | -.0254967 .0508094 -0.50 0.616 -.1250814 .0740879 _cons | -.3803795 .0544219 -6.99 0.000 -.4870445 -.2737145 ------------------------------------------------------------------------------

Informational content of the public debate

Estimates from propensity score matching

Treatment Outcome Propensity score

matching

Information about individual costs

YES

Information about PAYG sistem functioning

NO Attention

Information about deficit system

YES/NO

Summarising so far

• Citizens poorly informed

• Those who choose to be involved have the same characteristics of those more informed. Self-selection bias

• Attention could increase information about individual costs, less on iintergenerational redistribution and unsustainability

3. Information and opposition to reforms

• No majority in favour of reforms increasing sustainability

• Relevant cleavages:– Education– Age – Labour market status– Ideology

No reform gains a majority

Are you in favour of…(Italy 2004)

05

10152025303540

increasingretirement age

increasingcontribution rate

reducing pensionbenefits

Packaging is problematic

Number of reforms supported…(Italy 2004)

0

10

20

30

40

50

none 1 reform 2 reforms 3 reforms

Age divide is crucialItaly, Support to Unconditional O pting-O ut

(respondent as a percent of valid votes only)

57,3352,34

31,29

16-34 35-44 over 45

age classes

Germany, Support to Unconditional Opting-Out(respondents as a percent of valid votes only)

0

50

age classes

%

16-34 35-44 45-54 over 55

Favourable to reforms by awareness of individual costs

(which percentage of your salary is used to pay public pension

contributions?)

0

5

10

15

20

25

30

35

Increasing retirement age Increasing contribution rate Reducing pension benefits

Informed (knows her/his contribution rate) Not informed

Favourable to reforms by awareness of PAYG system functioning

(for which purpose are public pension contributions used for?)

0

5

10

15

20

25

30

35

Increasing retirement age Increasing contribution rate Reducing pension benefits

Informed (knows that contributions go to current pensioners) Not informed

Who is in favour of increasing the retirement age?

Probit estimates Number of obs = 627 LR chi2(7) = 33.37 Prob > chi2 = 0.0000 Log likelihood = -287.98822 Pseudo R2 = 0.0548 ------------------------------------------------------------------------------ refRetAge | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- male | .1457599 .1210145 1.20 0.228 -.0914241 .3829438 age56_65 | .3281721 .2091057 1.57 0.117 -.0816675 .7380117 university | .406008 .1396146 2.91 0.004 .1323684 .6796477 lowincome | .4682879 .201978 2.32 0.020 .0724183 .8641574 infopayg | -.0238039 .1194719 -0.20 0.842 -.2579647 .2103568 infocostind | .3902245 .1246733 3.13 0.002 .1458692 .6345798 infodeficit | .3081641 .139218 2.21 0.027 .0353019 .5810264 _cons | -1.48635 .1618994 -9.18 0.000 -1.803667 -1.169033 ------------------------------------------------------------------------------

There are also costs of information

In determining your retirement decision, which element will be more important?

(Italy, 2004)

0

10

20

30

40

50

personal reasons(not economic)

economic evaluation fear of new pensionreforms

Information does not reduce concerns

In determining your retirement decision, which element will be more important?

(by level of attention in following pension reform debate)Italy, 2004

0

10

20

30

40

50

high medium low none

personal reasons (not economic) economic evaluation

fear of new pension reforms

Press-media coverage may scare people

In determining your retirement decision, which element will be more important?

(Have you read newspaper articles/ watched Tv program concerning pension reform debate?)

Italy, 2004

0

10

20

30

40

50

yes no

personal reasons (noteconomic)

economic evaluation

fear of new pensionreforms

The “announcement effect”

Total numbers of workers who chose to retire, by year (1985-1998)

0

50,000

100,000

150,000

200,000

250,000

300,000

350,000

400,000

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Amato reform Dini reform Prodi reform

increase in the average number of retirements

Source: fRDB – CeRP calculations on LABOR – Inps data

Summarising

• Those more informed about costs and unsustainability support more reforms increasing sustainability

• Informed about PAYG more favourable to shrink size

• But is it due to self-selection or genuine information effects?

• Costs related to “informing” citizens: announcement (expectational) effects

Better ways to inform? The orange envelope