Relative Income Theory of Consumption:

An American economist J.S. Duesenberry put forward the theory of

consumer behaviour which lays stress on relative income of an
individual rather than his absolute income as a determinant of his
consumption. Another important departure made by Duesenberry
from Keynes’s consumption theory is that, according to him, the
consumption of a person does not depend on his current income but
on certain previously reached income level.

According to Duesenberry’s relative income hypothesis,

consumption of an individual is not the function of his absolute
income but of his relative position in the income distribution in a
society, that is, his consumption depends on his income relative to
the incomes of other individuals in the society. For example, if the
incomes of all individuals in a society increase by the same percent-
age, then his relative income would remain the same, though his
absolute income would have increased.

According to Duesenberry, because his relative income has

remained the same the individual will spend the same proportion of
his income on consumption as he was doing before the absolute
increase in his income. That is, his average propensity to consume
(APC) will remain the same despite the increase in his absolute
income.

As mentioned above, empirical studies based on time-series data

made by Kuznets reveal that over a long period the average
propensity to consume remains almost constant. Now,
Duesenberry’s relative income hypothesis suggests that in the long
run the community would continue to consume the same
proportion of income as its income increases.

According to Duesenberry, saving as a proportion of income of the

individuals with relatively low incomes would not rise much with
the increase in their incomes. That is, their savings would not rise to
the same proportion of income as was being done by the individuals
who had the same higher income prior to the present increase in
income.

This is because with the increase in incomes of all individuals by the

same proportion, the relative incomes of the individuals would not
change and therefore they would consume the same proportion of
their income. This applies to all individuals and households. It
therefore follows that assuming that relative distribution of income
remains the same with the growth of income of a society, its average
propensity to consume (APC) would remain constant.

Thus, this conclusion of the relative income hypothesis differs from

the Keynesian theory of consumption according to which, as seen
above, as absolute income of a community increases, it will devote a
smaller proportion of its income to consumption expenditure, that
is, its APC will decline.

It is important to note that relative income theory implies that with

the increase in income of a community, the relative distribution of
income remaining the same, does not move along the same
aggregate consumption function, but its consumption function
shifts upward. Since as income increases, movement along the same
consumption function curve implies a fall in average propensity to
consume, Duesenberry’s relative income hypothesis suggests that as
income increases consumption function curve shifts above so that
average propensity to consume remains constant.

This is illustrated in Figure 7.1. Suppose a family A’ has Y1 level of

income and is spending Y1A’ on consumption. Suppose its income
level rises to Y2. Now, its consumption would not rise only to Y2B
(i.e. equal to the consumption of the family B at Y2 income level) but
to Y2A’ where A’ lies on the same ray from the origin as the previous
point A of consumption. This implies that the consumption
expenditure of family A has risen in the same proportion as its
income with the result that its average propensity to consume
remains constant.

Likewise, if income of family B which is having consumption

expenditure Y2B at income level Y2, rises to Y3, its consumption
expenditure will increase to Y3B’ where B’ lies on the same ray from
the origin as B. This again means that the proportion of income
devoted to consumption by family B (i.e. its APC) remains constant
as there is increase in its absolute income.
Thus, if the proportion of income devoted to consumption of the
average family at each income level remains the same as its income
increases, the aggregate consumption of the community as
proportion of its income will also remain constant though its
absolute consumption and absolute savings will increase with the
absolute increase in income.

As income increases and a society moves along the same

consumption function curve, its average propensity to consume
falls. But Duesenberry’s relative income hypothesis suggests that as
income increases consumption function curver shifts above so that
average propensity to consume remains constant. In Figure 7.1 it
will be seen that if points A’ and B’ are joined together, we get, a
new consumption function curve C’C’.

Demonstration Effect:
By emphasising relative income as a determinant of consumption,
the relative income hypothesis suggests that individuals or
households try to imitate or copy the consumption levels of their
neighbours or other families in a particular community. This is
called demonstration effect or Duesenberry effect. Two things
follows from this. First, the average propensity to consume does not
fall.

This is because if incomes of all families increase in the same

proportion, distribution of relative incomes would remain
unchanged and therefore the proportion of consumption
expenditure to income which depends on relative income will
remain constant.
Secondly, a family with a given income would devote more of his
income to consumption if it is living in a community in which that
income is regarded as relatively low because of the working of
demonstration effect. On the other hand, a family will spend a lower
proportion of its income if it is living in a community in which that
income is considered as relatively high because demonstration
effect will not be present in this case.

For example, the recent studies of household expenditure made in

India reveal that the families with a given income, say Rs. 5000 per
month spend a larger proportion of their income on consumption if
they live in urban areas as compared to their counterparts in rural
areas.

The higher propensity to consume of families living in urban areas

is due to the working of demonstration effect where families with
relatively higher income reside whose higher consumption
standards tempt others in lower income brackets to consume more.

Ratchet Effect:
The other significant part of Duesenberry’s relative income
hypothesis is that it suggests that when income of individuals or
households falls, their consumption expenditure does not fall much.
This is often called a ratchet effect. This is because, according to
Duesenberry, the people try to maintain their consumption at the
highest level attained earlier. This is partly due to the demon-
stration effect explained above. People do not want to show to their
neighbours that they no longer afford to maintain their high
standard of living.
Further, this is also partly due to the fact that they become
accustomed to their previous higher level of consumption and it is
quite hard and difficult to reduce their consumption expenditure
when their income has fallen. They maintain their earlier con-
sumption level by reducing their savings. Therefore, the fall in their
income, as during the period of recession or depression, does not
result in decrease in consumption expenditure very much as one
would conclude from family budget studies.

This is illustrated in Figure 7.2 where on the X-axis we measure

disposable income and on the Y- axis the consumption and savings.
Starting with disposable income of zero, we assume that there is
steady growth of disposable income till it reaches Y1 .The linear
consumption function CLR is the long- run consumption function. It
will be seen from the figure that at Y1 level of disposable income, the
consumption expenditure equals Y1C1 .Now suppose with initial
income level Y1 there is recession in the economy with the result
that disposable income falls to the level Y0.
According to Duesenberry, consumption would not fall greatly to
the level Y0C0 as the long-run consumption function curve CLR would
suggest. In their bid to maintain their consumption level previously
reached people would now save less and reduce their consumption
level only slightly to Y0C’0 whereas point C’0 is on the short- run
consumption function curve CSR.
Since Y0C’0> Y0C0, the average propensity to consume at income
level Y0is greater at C’0 than at C1 at income level Y1 (A ray drawn
from the origin to the point C’0 will have greater slope than that of
OC1). When the economy recovers from recession and disposable
income increases, the economy would move along the short-run
consumption function curve CSR till the consumption level C1 is
reached at income level Y1. Beyond this, with the growth of income
the consumption will increase along the long-run consumption
function curve CLR.

Aggregate consumption function of the community:

From the analysis of demonstration and ratchet effects it follows
that Duesenb

This is illustrated in Figure 7.1. Suppose a family A’ has Y1 level of

income and is spending Y1A’ on consumption. Suppose its income
level rises to Y2. Now, its consumption would not rise only to Y2B
(i.e. equal to the consumption of the family B at Y2 income level) but
to Y2A’ where A’ lies on the same ray from the origin as the previous
point A of consumption. This implies that the consumption
expenditure of family A has risen in the same proportion as its
income with the result that its average propensity to consume
remains constant.
Likewise, if income of family B which is having consumption
expenditure Y2B at income level Y2, rises to Y3, its consumption
expenditure will increase to Y3B’ where B’ lies on the same ray from
the origin as B. This again means that the proportion of income
devoted to consumption by family B (i.e. its APC) remains constant
as there is increase in its absolute income.
Thus, if the proportion of income devoted to consumption of the
average family at each income level remains the same as its income
increases, the aggregate consumption of the community as
proportion of its income will also remain constant though its
absolute consumption and absolute savings will increase with the
absolute increase in income.

As income increases and a society moves along the same

consumption function curve, its average propensity to consume
falls. But Duesenberry’s relative income hypothesis suggests that as
income increases consumption function curver shifts above so that
average propensity to consume remains constant. In Figure 7.1 it
will be seen that if points A’ and B’ are joined together, we get, a
new consumption function curve C’C’.
Demonstration Effect:
By emphasising relative income as a determinant of consumption,
the relative income hypothesis suggests that individuals or
households try to imitate or copy the consumption levels of their
neighbours or other families in a particular community. This is
called demonstration effect or Duesenberry effect. Two things
follows from this. First, the average propensity to consume does not
fall.

This is because if incomes of all families increase in the same

proportion, distribution of relative incomes would remain
unchanged and therefore the proportion of consumption
expenditure to income which depends on relative income will
remain constant.

Secondly, a family with a given income would devote more of his

income to consumption if it is living in a community in which that
income is regarded as relatively low because of the working of
demonstration effect. On the other hand, a family will spend a lower
proportion of its income if it is living in a community in which that
income is considered as relatively high because demonstration
effect will not be present in this case.

For example, the recent studies of household expenditure made in

India reveal that the families with a given income, say Rs. 5000 per
month spend a larger proportion of their income on consumption if
they live in urban areas as compared to their counterparts in rural
areas.
The higher propensity to consume of families living in urban areas
is due to the working of demonstration effect where families with
relatively higher income reside whose higher consumption
standards tempt others in lower income brackets to consume more.

Ratchet Effect:
The other significant part of Duesenberry’s relative income
hypothesis is that it suggests that when income of individuals or
households falls, their consumption expenditure does not fall much.
This is often called a ratchet effect. This is because, according to
Duesenberry, the people try to maintain their consumption at the
highest level attained earlier. This is partly due to the demon-
stration effect explained above. People do not want to show to their
neighbours that they no longer afford to maintain their high
standard of living.

Further, this is also partly due to the fact that they become
accustomed to their previous higher level of consumption and it is
quite hard and difficult to reduce their consumption expenditure
when their income has fallen. They maintain their earlier con-
sumption level by reducing their savings. Therefore, the fall in their
income, as during the period of recession or depression, does not
result in decrease in consumption expenditure very much as one
would conclude from family budget studies.

This is illustrated in Figure 7.2 where on the X-axis we measure

disposable income and on the Y- axis the consumption and savings.
Starting with disposable income of zero, we assume that there is
steady growth of disposable income till it reaches Y1 .The linear
consumption function CLR is the long- run consumption function. It
will be seen from the figure that at Y1 level of disposable income, the
consumption expenditure equals Y1C1 .Now suppose with initial
income level Y1 there is recession in the economy with the result
that disposable income falls to the level Y0.
According to Duesenberry, consumption would not fall greatly to
the level Y0C0 as the long-run consumption function curve CLR would
suggest. In their bid to maintain their consumption level previously
reached people would now save less and reduce their consumption
level only slightly to Y0C’0 whereas point C’0 is on the short- run
consumption function curve CSR.
Since Y0C’0> Y0C0, the average propensity to consume at income
level Y0is greater at C’0 than at C1 at income level Y1 (A ray drawn
from the origin to the point C’0 will have greater slope than that of
OC1). When the economy recovers from recession and disposable
income increases, the economy would move along the short-run
consumption function curve CSR till the consumption level C1 is
reached at income level Y1. Beyond this, with the growth of income
the consumption will increase along the long-run consumption
function curve CLR.

Aggregate consumption function of the community:

From the analysis of demonstration and ratchet effects it follows
that Duesenberry’s relative income hypothesis provides an
explanation for why aggregate consumption function of the
community may be flatter than the family budget studies would
suggest. Duesenberry emphasizes that it is relative income rather
than absolute income which determines consumption expenditure
of households.

When income of the community increases, relative income

remaining constant, the proportion of consumption expenditure to
income will not increase much because relative incomes of the
households remain the same (Note that this implies that saving
ratio will not rise much).

Due to demonstration effect every household will increase its

expenditure in the same proportion as the increase in income. On
the other hand, if the income of the community decreases, the
consumption expenditure would not decline much due to the
ratchet effect according to which people try to maintain their
previously attained higher level of consumption. This makes the
consumption function of the community flatter than suggested by
the cross-sectional family budget studies.

Further, it also follows from the Duesenberry relative income

hypothesis that short-run aggregate consumption function of the
community is linear rather than curved. As stated above, if, in the
short run, the level of income increases, the proportion of
consumption expenditure to income is not likely to increase much
due to the operation of demonstration effect and with the fall in
income the proportion of consumption to income is not likely to
decline much due to the ratchet effect.

This makes the short-run aggregate consumption function of the

community linear. It is worth noting that Duesenberry’s theory
assumes that relative distribution of income does not change much.
This is in accord with the facts of the real world situation where
changes in income distribution do not take place in the short run.
Thus Duesenberry’s theory provides a convincing explanation in
terms of demonstration and ratchet effects why aggregate
consumption function is linear rather than nonlinear.