Public Finance, 2nd Edition

M

h

t

evaluated in this way standardizes both for changes in prices since 1973

and for the needs of families with characteristics diVerent from A

R

. The

number of poor equals the number of families for which M

h

t

< M

Z

. Alterna-

tively, the number of poor equals the number of families for which V

h

t

< V

R

Z

,

the poverty line level of utility, which is constant over time.

Slesnick found that his consumption-based poverty count was lower than

the oYcial poverty count in all but 4 years of the sample period and was

substantially lower by the end of the period. From 1981 through 1989, Sles-

nick's estimated poverty rate was approximately four percentage points below

the oYcial poverty rate each year. For example, in 1989 Slesnick estimated

that 8.4% of all families were poor, whereas the oYcial poverty rate was 12.8%.

The consumption-based poverty rate is below the oYcial income poverty

rate primarily for three reasons. One is that 40% of the poor own their own

homes, so that they consume a fairly large amount of capital services; capital

service Xows account for 10 to 13% of the total expenditures of the poor in

Slesnick's sample. The second is that a large number of the poor dissave; their

incomes are temporarily low and they dissave to maintain their standard of

living. The third factor that drove the Slesnick poverty counts down sharply

in the 1980s was a change in family characteristics that helped to move

families out of poverty, both directly and indirectly by its eVect on the

composition of family expenditures

Slesnick argues that his consumption-based poverty count is superior to

the oYcial count because it better reXects families' permanent economic

situations. The poor by his measure are more likely to be permanent income

poor than the ``oYcial'' current income poor. He found that budgets of the

consumption poor contain a lower percentage of capital services than

the consumption nonpoor because they are less likely to own their own

homes. He also found that the consumption poor devote a higher percentage

of their budgets to purchases of food.

SOCIAL WELFARE AND SOCIAL MOBILITY

Bergson and Samuelson conceived of their individualistic social welfare func-

tion in terms of end-results equity, as a device for evaluating the ethical content

of social outcomes. All our applications of the social welfare function so far

have been in this vein. Despite its end-results orientation, economists have also

used the social welfare function to measure the ethical implications of one

common measure of process equity, the degree of social mobility in society.

Social mobility refers to the ability of individuals (families) to move

throughout the distribution of income over time. It is closely related to the

other widely held notion of process equity, equal opportunity. At one ex-

treme is the caste system, a completely immobile society. People are assigned

a position in the distribution at birth and can never move; there is no

134 SOCIAL WELFARE AND SOCIAL MOBILITY