The Poverty & Inequality Effects of Pensions Dr Michel Collins, - PowerPoint PPT Presentation

The Poverty & Inequality Effects of Pensions Dr Micheál Collins, UCD ml.collins@ucd.ie NERI Labour Market Conference, September 2020

Outline 1. Research Question 2. Why? 3. Data and Methods 4. Initial Results 5. Next Steps… 2

1. Research Question  A core objective of pensions is poverty avoidance  A core objective of redistribution is poverty avoidance  Less direct focus on inequality effects  implicit in poverty aims…  Q:  How effective are Ireland’s existing pension policy tools at reducing poverty and inequality? 3

1. Research Question  A core objective of pensions is poverty avoidance  A core objective of redistribution is poverty avoidance  Less direct focus on inequality effects  implicit in poverty aims…  Q:  How effective are Ireland’s existing pension policy tools at reducing poverty and inequality? 4

2. Why?  An interest in the inequality effects of taxation measures  interesting UK work from NSO…  Broadened to other wing of redistribution:  welfare effects  within this state pensions  expand to include private pensions…  A pensions detour… 5

3. Data and Methods  For this initial consideration:  CSO SILC 2017  Microdata from the ISSDA at UCD  5,029 households and 12,612 individuals  nationally representative, detailed income data, detailed socio- economic characteristics…  linked to DEASP and Revenue sources to verify much of the income information  Collins and Hughes (earlier years) showed robustness for pensions analysis 6

 Pensions tools available to be examined: Work related pensions  Private pensions (occupational and personal)  State pensions occupational Social welfare pensions  State pension social welfare  OAP  survivors pension 7

 Inequality Decomposition  Gini coefficient  from 0 to 100  isolate stand-alone impact of each policy measure on the Gini  Formally: Reynolds-Smolensky index  Impact = Gini with – Gini without  e.g. if Gini is 50 but increases to 55 if you remove the effect of redistribution by measure X, then the stand-alone effect of policy measure X is that it decreases the Gini by 5% 8

 Poverty Decomposition  Similar approach, looking with and without pension tool  Using:  Poverty risk = % of population below 60% median income poverty line (equivalised)  Poverty gap = average distance below the poverty line  Poverty count = number of people in poverty 9

4. Initial Results  Initial…  Baseline positions:  results from SILC 2017 published report  results with no pensions  The simulations to identify effects 10

Table 1: The Poverty and Inequality Effect of Pensions, 2017 (ceteris paribus) Mean Poverty Gap € per Poverty Gini Poverty Count week Baseline 2017 SILC 15.7 31.5 755,592 56.35 Baseline no pensions 28.8 40.0 1,384,955 125.11

Table 1: The Poverty and Inequality Effect of Pensions, 2017 (ceteris paribus) Mean Poverty Gap € per Poverty Gini Poverty Count week Baseline 2017 SILC 15.7 31.5 755,592 56.35 Baseline no pensions 28.8 40.0 1,384,955 125.11 Simulations – results are changes to the no pensions baseline Pensions overall -13.1 -8.5 -629,363 -68.76

Table 1: The Poverty and Inequality Effect of Pensions, 2017 (ceteris paribus) Mean Poverty Gap € per Poverty Gini Poverty Count week Baseline 2017 SILC 15.7 31.5 755,592 56.35 Baseline no pensions 28.8 40.0 1,384,955 125.11 Simulations – results are changes to the no pensions baseline Pensions overall -13.1 -8.5 -629,363 -68.76 Work related pensions -3.5 -1.7 -164,437 -15.48

Table 1: The Poverty and Inequality Effect of Pensions, 2017 (ceteris paribus) Mean Poverty Gap € per Poverty Gini Poverty Count week Baseline 2017 SILC 15.7 31.5 755,592 56.35 Baseline no pensions 28.8 40.0 1,384,955 125.11 Simulations – results are changes to the no pensions baseline Pensions overall -13.1 -8.5 -629,363 -68.76 Work related pensions -3.5 -1.7 -164,437 -15.48 occupational state pensions -3.1 -1.6 -146,419 -14.73 private pensions -0.3 -0.1 -12,028 -0.76

Table 1: The Poverty and Inequality Effect of Pensions, 2017 (ceteris paribus) Mean Poverty Gap € per Poverty Gini Poverty Count week Baseline 2017 SILC 15.7 31.5 755,592 56.35 Baseline no pensions 28.8 40.0 1,384,955 125.11 Simulations – results are changes to the no pensions baseline Pensions overall -13.1 -8.5 -629,363 -68.76 Work related pensions -3.5 -1.7 -164,437 -15.48 occupational state pensions -3.1 -1.6 -146,419 -14.73 private pensions -0.3 -0.1 -12,028 -0.76 Social Welfare pensions -9.1 -5.0 -435,255 -43.44

Table 1: The Poverty and Inequality Effect of Pensions, 2017 (ceteris paribus) Mean Poverty Gap € per Poverty Gini Poverty Count week Baseline 2017 SILC 15.7 31.5 755,592 56.35 Baseline no pensions 28.8 40.0 1,384,955 125.11 Simulations – results are changes to the no pensions baseline Pensions overall -13.1 -8.5 -629,363 -68.76 Work related pensions -3.5 -1.7 -164,437 -15.48 occupational state pensions -3.1 -1.6 -146,419 -14.73 private pensions -0.3 -0.1 -12,028 -0.76 Social Welfare pensions -9.1 -5.0 -435,255 -43.44 old age pensions (cash) -8.8 -4.8 -420,648 -42.11 survivors benefit (cash) -0.3 -0.2 -14,607 -1.33

5. Next Steps  Suggestions?  Plan for:  multiple years of SILC (to establish the average effect, although will do detailed analysis on just one year)  will add 2018 data (new!)  split poverty indicators to examine above and below 65yrs / 70yrs  gender differences  A later in 2020 project… 17

The Poverty & Inequality Effects of Pensions Dr Micheál Collins, UCD ml.collins@ucd.ie NERI Labour Market Conference, September 2020