The Average Propensity to Consume Out of Full Wealth: Testing a New Measure Laurie Pounder.

The Average Propensity to Consume Out of Full Wealth:

Testing a New Measure

Laurie Pounder

Full Wealth: The Right Measure of Wealth for Consumption

Lifecycle/PIH theory since Modigliani says consumption should depend on all current and future resources (including financial and human wealth.) • Essentially a stock value of permanent income

from today forward• I call this PDV of all resources:

“Modigliani full wealth” = M

Unprecedented Ability to Measure Full WealthHealth and Retirement Study

Expected present value of resources:

M = Net Worth + Human Wealth• Net Worth = 10 categories of assets less 3 categories

of debt• Human Wealth=

Earnings+Pensions+Social Security+Other Transfers

(deterministic for older households)

Full Wealth is Not Just Scaled-Up Net Worth

Age Profile of Wealth

20

0,0

00

40

0,0

00

60

0,0

00

80

0,0

00

10

0000

0

55 60 65 70 75Age of Household Head

Full Wealth Cash-on-Hand Net Worth

Full Wealth

Net worth

Full Wealth Has Less Variance…

Coefficients of Variation

CV Mean

Full Wealth 0.99 $738,100

Net Worth 1.68 $324,300

Income 1.24 $62,100

Consumption 0.76 $40,300

…and is more equally distributed0

.2.4

.6.8

1

0 .2 .4 .6 .8 1Cumulative population proportion

Lorenz Curve Full Wealth Lorenz Curve Net Worth

Full Wealth

Net worth

Lorenz Curves

The Average Propensity to Consume Out of Full Wealth

Lifecycle model:

• Very limited source of variation in C/M across households

• C/M changes only slowly over time (from mortality, changes in returns expectations, or changes in preferences)

• C/M does not change with income shocks if consumption responds quickly

Which Implies…

Relative to C/Income or C/NetWorth, C/M Should Have:

• Lower variance• Higher covariance over time• Lower correlation with “circumstances” such as:

– Income Profile• Having a pension or the generosity of pension and social security

benefits (income replacement rate in retirement) • Earnings profile over lifetime

– Past Income Shocks

Also ∆(C/M) Should Have:• Lower correlation with past shocks both to income and

to full wealth

And the data says…

Std. Dev. Mean Median CV

C/M 2001 .058 .078 .060 0.74

C/M 2003 .062 .084 .067 0.74

C/NW 2001 3.26 1.05 .221 3.10

C/NW 2003 12.56 2.59 .256 4.85

C/I 2001 1.47 1.22 .828 1.20

C/I 2003 1.37 1.24 .888 1.10

Lower and more consistent variance

Covariance 2001&2003

C/M 0.70

C/NW 0.37

C/I 0.27

And higher covariance over time

Circumstances

• Traditional savings or consumption rates (C/I) have “noise” from circumstances, both cross-sectionally and longitudinally

• Examples:– Households expecting generous DB pension

income will save less than otherwise identical households with little or no DB pension

– Households experiencing a temporary positive income shock will save more that period

Lifecycle Model Illustrations

Standard Lifecycle Consumption

Age

Consumption Income Net Worth

Rates of Consumption Out of Alternate Measures of Resources: Income and Full Wealth

20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65 68 71 74 77 80

Age

Rat

e of

C

onsu

mpt

ion

C/Income C/Full Wealth

Comparison of Baseline to Household with Lower Retirement Income

Income and Net Worth with Different Retirement Incomes

Income Net Worth Baseline Income Baseline Net Worth

`

Rate of Consumption Out of Income

C/In

com

e

Lower Retirement Baseline

Consumption Rate Out of Full Wealth with Different Retirement Incomes

Age

Rat

e of

C

onsu

mpt

ion

Lower Retirement Baseline

Income Shocks

Standard Lifecycle withTemporary Income Shock

Age

Consumption Income Net Worth

Comparison of Baseline and Shocked Household

Consumption Rate Out of Income with Income Shock

Age

Rat

e o

f C

on

sum

pti

on

C/I With Shock C/I Baseline

Consumption Rate Out of Full Wealth with Income Shock

AgeC/

M

C/M with Shock C/M Baseline

Testing Circumstances

Circumstance: Generosity of retirement benefits (DB pension and Social Security)

Measure:

RetRatio= Ratio of PV(Pension+Social Security) to Average Earnings Over Ages 45-55

Outcome: C/M is less correlated

Retirement/Earnings Ratio

Bivariate OLS Coefficient & T-stat R2

std(C/M) 2001 on RetRatio 0.003 (1.2) 0.00

std(C/NW) 2001 on RetRatio 0.013*** (6.0) 0.03

std(C/I) 2001 on RetRatio 0.005** (2.1) 0.01

std(C/M) 2003 on RetRatio -0.001 (-0.4) 0.00

std(C/NW) 2003 on RetRatio 0.016*** (4.8) 0.03

std(C/I) 2003 on RetRatio 0.007** (2.3) 0.01

Coefficients represent fraction of standard deviation from mean so can be compared across dependent variables

Income Shocks

Circumstance: Past Income Shock

Measure: Change in Earnings over previous years

Outcome With Levels: results mixed: C/M less correlated than C/I in 2001; less correlated for large shocks in 2003

Income Shocks on Levels of Consumption Rates

2001 Dependent Variable→ std(C/M) std(C/NW) std(C/I)

Independent Variables↓

Y Shock 2000-2001 0.163** (2.0) 0.124* (1.6) -0.263***(-3.2)

Y Shock 1999-2000 -0.064 (-0.7) -0.077 (-0.8) -0.314***(-3.3)

2003 std(C/M) std(C/NW) std(C/I)

Y Shock 2001-2003 -0.003 (0.6) 0.004 (0.4) -0.008 (-1.3)

Y Shock 2000-2001 0.135 (1.4) -0.079 (-0.8) 0.094 (1.0)

2003 with Large Shocks

2003 std(C/M) std(C/NW) std(C/I)

>25% Negative

Y Shock 2001-2003

-0.103

(-0.9)

-0.314***

(-2.7)

0.641***

(6.1)

>25% Positive

Y Shock 2001-2003

0.085

(0.9)

0.161*

(1.7)

-0.390***

(-4.4)

>25% Negative

Y Shock 2000-2001

0.028

(0.2)

-0.310**

(-2.3)

0.311**

(2.5)

>25% Positive

Y Shock 2000-2001

0.126

(1.0)

-0.073

(-0.6)

-0.020

(-0.2)

Change in C/M Less Correlated With Shocks

Dependent Variable→ ∆(C/M) ∆(C/NW) ∆(C/I)

Independent Variables↓

Y Shock 2000-2001 -.050 (-1.3) -.175** (-2.0) .400*** (6.2)

Y Shock 1999-2000 .001 (0.0) .044 (0.4) .168** (2.4)

Changes in M

• Since M is an expected value of current and future resources, any change in M must be unexpected, unlike a change in income

• If consumers adjust relatively quickly to changes in M, then C/M should be relatively invariant to such changes

• Instrument for change in M: Unexpected retirement

Change in C/M Less Affected by Unexpected Changes in M

Dependent Variable→ ∆(C/M) ∆(C/NW) ∆(C/I)

Independent Variable↓

Unexpected Retirement between 2001 & 2003

0.003

(0.4)

-0.077

(-1.0)

0.267***

(3.5)

R2 0.00 0.01 0.02

Conclusion

• Empirically, full wealth, M, and C/M match expected distribution characteristics

• The level of C/M has less correlation with tested circumstances than either C/NW or C/I

• The change in C/M is relatively invariant to recent income and employment shocks and changes in M when compared to C/NW or C/I