Poverty and household spending in Britain

1
Poverty and household spending in Britain

• Mike Brewer
• Alissa Goodman
• Andrew Leicester
• Institute for Fiscal Studies
• 17th May 2006

2
The effect of increased benefit entitlements on
pensioners spending Mike Brewer
3
Motivation

• State benefits for 60 risen under Labour, yet
  spending poverty of pensioners little changed
• Have extra benefits improved pensioners living
  standards?
• Related work
• Meyer and Sullivan (2004) US data, lone
  parents
• Gregg, Waldfogel and Washbrook (2004, 2006) UK
  data, families with children
• Blow, Walker and Zhu (2005), UK data, families
  with children
• Munro, Walker and Zhu (ongoing) UK data, winter
  fuel allowance

4
Outline

• Policy changes affecting pensioners
• Method and data
• Results
• Conclusions

5
Main benefit changes affecting pensioners since
1997

• Rise in basic state pension (April 2001 2002)
• Increases in means-tested benefits (from April
  1999) and introduction of pension credit (from
  2003)
• Equalisation of pensioner premia in means-tested
  benefits (by 2001)
• Winter fuel allowance (from 1999)

6
Changes to benefit entitlement for single
pensioners (19961)
Graph shows maximum entitlement to IS or BSP for
single pensioner
7
Overview of method

• Compare (changes in) spending of pensioners
  affected by rise in benefits to pensioners not
  affected
• Attribute difference to policy.
• Called conditional difference-in-differences.

8
Difference-in-differences data

• FES/EFS from 1996/7 to 2002/3
• Single adults born before April 1936
• Aged 60 in 1996, 66 in 2002 (pseudo-panel)
• 3,056 poor pensioners (entitled to a
  means-tested benefit under 1996/7 system)
• 1,281 young 1,775 old
• 1,778 rich pensioners (not entitled to a
  means-tested benefit under 2002/3 system)
• Some pensioners omitted entirely (neither poor
  nor rich)

9
Changes in benefit entitlements, income and
spending, 1996/7-2002/3
10
Difference-in-differences overview

• Compare spending before and after rise in
  means-tested benefits (April 1999)
• Rich pensioners tell us about general trends
  affecting pensioners B-A
• Poor pensioners tell us about general trends and
  impact of policy D-C.
• Difference tells us about impact of policy (D-C)
  (B-A)
• Assumes common trends
• Control for various factors (regression-adjusted
  DiD)
• Also compare young and old poor pensioners

Mean spend 4/96 to 3/99 4/99 to 3/03
Rich A B
Poor C D
11
Impact of benefit changes on pensioners
Impact of policy on log( . ) April 1999 MIG (Poor vs Rich) April 2001 eqn (Young vs Old)
Income 0.112 0.016
Spending (non-housing) 0.096 0.097
Spending on basics 0.030 -0.008
Spending on non-basics 0.149 0.184
significant _at_ 10 significant _at_ 1
12
Impact of benefit changes on pensioners
Impact of policy on log( . ) April 1999 MIG (Poor vs Rich) April 2001 eqn (Young vs Old)
Income 0.112 0.016
Spending (non-housing) 0.096 0.097
Spending on basics 0.030 -0.008
Spending on non-basics 0.149 0.184
significant _at_ 10 significant _at_ 1
13
Impact of benefit changes on pensioners
Impact of policy on log( . ) April 1999 MIG (Poor vs Rich) April 2001 eqn (Young vs Old)
Income 0.112 0.016
Spending (non-housing) 0.096 0.097
Spending on basics 0.030 -0.008
Spending on non-basics 0.149 0.184
significant _at_ 10 significant _at_ 1
14
Conclusions

• Pensioners look poorer when assessed using
  spending than income
• Recent rises in means-tested benefit for
  pensioners were translated into higher spending
• Results rely on untested common trends
  assumption evidence stronger for introduction of
  MIG than equalisation of age-related premia

15
Summing up

• Living standards have risen whether measured by
  income or spending
• Increased expenditure poverty rate since 1997
  even as income poverty declined
• Reasons for different trends not yet clear
• Recent rises in means-tested benefit for
  pensioners were translated into higher spending