Expenditure and Consumption
James Banks (UCL, IFS)
Andrew Leicester (IFS)
Motivation


Utility / welfare depends on consumption rather than
income
Particularly important distinction for the elderly:
Permanent Income Hypothesis – spending above income?
 Or spending below income?


Uncertain lifespan / healthcare needs
 Bequests


Retirement-Savings Puzzle
Key policy interest

Pensioner spending poverty higher than income poverty
 Fuel poverty

Broad range of welfare indicators in ELSA
Measurement

Spending not consumption


Information on durable ownership / replacement / expenditures
Selected sample





Each benefit unit contains an eligible ELSA sample member
Full reporting of all expenditure items
6,557 people (2,908 males, 3,649 females) in 4,295 households
Adjust for inflation and household composition
No weighting
Spending on what?

Food In
Clothing
Fuel

Food Out




Key ‘necessities’
Restaurant meals, takeaway, canteens
Leisure

e.g. cinema, sport, fees/subscriptions, internet/cable, TV licence

Cash transfers

Incomplete measure of total spending
Expenditure by age
Food in
Fuel
Clothing
Food out
Leisure
Transfers
125
£ per week
100
75
50
25
0
52-54
55-59
60-64
65-69
70-74
75-79
80+
All
Necessities budget share by age
Food in
Fuel
Clothing
40
Per cent
30
20
10
0
52-54
55-59
60-64
65-69
70-74
75-79
80+
All
Necessities budget share by age / wealth
Poorest
2nd quintile
3rd quintile
4th quintile
Richest
40
Per cent
30
20
10
0
52-59
60-74
75 and over
All
Necessities budget share by age / wealth
Poorest
2nd quintile
3rd quintile
4th quintile
Richest
40
Per cent
30
20
10
0
52-59
60-74
75 and over
All
Necessities budget share by age / wealth
Poorest
2nd quintile
3rd quintile
4th quintile
Richest
40
Per cent
30
20
10
0
52-59
60-74
75 and over
All
Proportion of people making transfers, by
age / wealth
3rd quintile
2nd quintile
Poorest
4th quintile
Richest
90
23
80
19
70
Per cent
11 18
9
19 26
17 33
14
15
11
12
60
50
24
25
10
9
9
11
9
40
30
20
10
0
52-59
60-74
75 and over
All
Proportion of people making transfers, by
age / wealth
3rd quintile
2nd quintile
Poorest
4th quintile
Richest
90
23
80
19
70
Per cent
11 18
9
19 26
17 33
14
15
11
12
60
50
24
25
10
9
9
11
9
40
30
20
10
0
52-59
60-74
75 and over
All
0
Dishwasher
Digital TV
DVD
Computer
Dryer
CD
M'wave
VCR
Washing
Machine
Freezer
Landline
TV
Per cent
Consumer durable ownership rates
100
80
60
40
20
Consumer durables



Younger sample members more likely to own
Little difference by wealth for TV, video, freezer, washing
machine, microwave, landline phone, DVD player
High replacement / purchase rates during previous 2 years for
recent innovations or goods where technology changes

Television (23%)
 Computer (23%)
 Digital TV (15%)
 DVD player (44%)


Replacement rate invariant to wealth
Wealthier households buy more expensive models
Digital TV adoption, by age/sex
Men
Women
70
60
Per cent
50
40
30
20
10
0
52-54
55-59
60-64
65-69
70-74
75-79
80+
All
“Fuel Poverty”, by age / sex
Men
Women
16
14
Per cent
12
10
8
6
4
2
0
52-54
55-59
60-64
65-69
70-74
75-79
80+
All
“Fuel Poverty”, by age / CASP-19 tertile
Lowest
Middle
Highest
16
14
Per cent
12
10
8
6
4
2
0
52-59
60-74
75 and over
All
“Fuel Poverty”, by age / CASP-19 tertile
Lowest
Middle
Highest
16
14
Per cent
12
10
8
6
4
2
0
52-59
60-74
75 and over
All
Putting It Together …

What are the key factors that influence spending? Need
multivariate analysis …

Lots of interesting questions we can ask

Is there an independent role for wealth once we control for health?
 Does age matter once we control for labour market status?
Low-spending pensioner households

Why do pensioners spend less than their income:

Uncertain longevity?

Simple model controlling for wealth, health, age
 Pensioners who say they expect to live until old age spend same share
of their income than those who say they are unlikely
 1% increase in spending raises expectation of living to age 80 by more
than 2 percentage points

Bequest motive?


Probability of leaving a bequest of £50,000 + has no independent effect
on share of income spent
Both very early findings but interesting (and puzzling!)
Conclusions







Expenditure / durables provide insight into economic well-being
‘Luxury’ spending items diminish strongly with age, ‘necessities’
less so
Transfers to others an important part of the budget of many
elderly households
Spending positively correlated with other dimensions of wellbeing
Danger of older / poorer pensioners being left behind in terms of
new technologies
Much interesting work that can be done!
Key interest in collecting longitudinal data on spending:
changing patterns around retirement and as pensioners age
Expenditure and Consumption
James Banks (UCL, IFS)
Andrew Leicester (IFS)