This document discusses the challenges of an aging society and pension reform. It provides background on the speaker and their research analyzing pension systems using economic models. Experience from reforms in Poland showed that while defined contribution pensions balanced, other benefits led to deficits without further reform. Modeling indicates that reducing retirement ages is harmful and that most people are not fully rational in savings decisions. Effective policies are needed to encourage voluntary pension savings given issues around incomplete rationality and uncertainty around longevity, returns, and system changes. There are still many unknowns around interpreting savings behaviors and balancing responsibility with redistribution in pension systems.

1. Challenges of an Aging Society
LSE Global Pensions Programme 2019
Joanna Tyrowicz

2. Who am I and why am I talking to you?
o GRAPE is an economic (NGO) research center
o Pension program at GRAPE
o Research (pure & basic: journal publications, PhD theses, etc.)
o Policy: European Commission, OECD, The World Bank
o EU-China High Level Round Table: reforming pensions
o TFI (ING, CEPR, DELL, DELOITTE): study societal challenges
o I am a professor of economics
o Labor and pension issues
o Fulbright (Columbia University)
o Economic Advisor at National Bank of Poland (2009-2017)
o I work in Poland and in Germany
o Cofounding GRAPE + co-running the pension program at GRAPE

3. o Rational people plan ahead
o Humans are not at all wired to plan
(see Dan Gilbert, Harvard; Daniel Kahnemann, Princeton)
o Moreover, the problem is very complicated
o Planning under uncertainty per se
o Individual longevity also unknown
o Biases in understanding interest (and compounding)
o Can policies help? Pension systems!
How good are we in providing for old-age?

4. Experience from Poland (and the region)
o After the 1990s reforms, pension systems were to be balanced
o Absent reform, pension system would have generated deficit of ~8%
of GDP each year
o After reform, defined contributions are balanced, but
o minimum pension benefit guarantee (4-5% GDP)
o inappropriate use of actuarial tables (1.5% GDP)
o inconsistent inheritance rules (~0.7% GDP)
o Outcome: low pensions, with low indexation and mounting deficit
o Private voluntary pension savings needed more than ever

5. Analyzes of GRAPE (so far)
o State-of-the-art overlapping generations models (ex ante!)
o Full demographic forecast
o People adapt to changes in taxes, contributions, interest rates…
o People are heterogeneous: preferences and biases
o What do we know so far?
o The reform from 1999 was ok.
o The changes from 2011/2013 not so much …
o … but could not be prevented
o Reducing the retirement age is a very bad idea
o Et cetera
o Can we effectively encourage to raising voluntary pension savings?

6. How do we run such analyses?
o People optimize “happiness” during lifetime
o They like consumption, but do not like to work so much
o They have expectations about longevity, taxes, pensions, wages, interest
rates …
o Firms hire work and use savings as productive capital
o The government
o Collects taxes, pays govn’t consumption, pays debt interest …
o … and balances pensions (collects contributions and pays benefits)
+ Capital pillar in the system: mandatory or voluntary
We calibrate such a model to replicate a true economy …
… and then we play!

7. Reducing retirement age is a very bad idea
0%
10%
20%
30%
40%
50%
60%
70%
80%
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
Year of birth
% of pension recipients
whose pension falls short of the minimum benefit
retirement age of 67
retirement age of 60/65

8. A fully rational person would accumulate …
0
1
2
3
4
5
6
7
8
21
24
27
30
33
36
39
42
45
48
51
54
57
60
63
66
69
72
75
78
81
84
87
90
93
96
99
Totalassets
Age
No pension system
No pension system + longevity
DC system + longevity

9. … which requires the following savings rate
0
0.1
0.2
0.3
0.4
0.5
0.6
21222324252627282930313233343536373839404142434445464748495051525354555657585960
Savingsrate
Age

10. Issue: how many of us are fully rational?
o Typically no more than 13%
o Testing in one domain does not need to translate to other domains
o preferences and (rational) inattention
o „correct” choices despite bad premises?
o Many ways to be incompletely rational
o present bias
o lacking financial literacy
o failing to (fully) internalize changes
o …

11. Incomplete rationality

12. Instead of summary
o Plenty of known unknowns:
o How to interpret low savings/participation in savings vehicles?
o How to interpret low annuitization?
o Role of trust in the system
o Compulsory vs voluntary
o Some additional unknowns are value based
o How much redistribution vs how much responsibility
o Long-run effects of longevity

13. Thank you for your attention
w | grape.org.pl Joanna Tyrowicz: j.tyrowicz@grape.org.pl
fb | GRAPE.org
tt | GRAPE_ORG