Consumption is a function of income and it is the largest component in the expenditure function.
1946 Kuznets: findings over time regarding C. Considere three features:
1. as Y increases, when c.f. acrss hsehlds/income groups, the richer hsehlds had lower apc.
2.During slump, high apc bec. Y is lower. During boom, Y higher than normal, apc falls.
3.In the LR, C to Y ratio or apc = constant. (75% of national income).
KEYNESIAN CONSUMPTION FUNCTION
*******************************
see seperate sheet
THE LIFE CYCLE APPROACH
***************************
-views individuals as planning their consumption and savings behaviour over long periods, with the intention of allocating their consumption in the best possible way over thier entire lifetimes.
-views savings resulting from desire to provide for consumption in old age.
-Consumption a constant proportion of individuals permanent income.
-Given a consumption function C= a WR + c YL
where WR = real wealth
a = mpc out of wealth
YL = labour income
-Consumption over lifetime cannot exceed lifetime income unless born with wealth. We initially assume this not to be the case. It is id. that part of the consumers problem is finding a limit on lifetime consumption.
-Theory of consumer behavior starts with utility max/min - referred to as OPTIMISATION.
Diffc. arises in that decisions to consume are not made in one snapshot period of time. e.g. purchasing durable goods - car - consume overmore than 1 unit of time. Assume want to distribute consumption over lifetime so have a smooth or flat consumption pattern.
The optimising individual who thinks of consumption over time is described as an INTER-TEMPORAL OPTIMISER.
-Individual confronmted with 2 decisions:
1.increase current consumption by borrowing against future consumption.
2.save and consume more in the next period.
It all depends on - indiv. taste reflected by Indifference curve.
- income budget constraint
-interest rate slope of budget constraint
Rather than have high income in one period and low in the next, the preferred profile is to have a smooth consumption pattern. Why?
Due to the rule of dim. marginal utility.
Consider: a period of plenty and a period of scarcity. The loss in marginal utility from the period of penty to the period of scarcity is more than justified by the gain in utility in the period of scarcity.
-Income fluctuates in an individuals lifetime. Low during the early years, high during older years. Kowing his tastes and endowments he can smooth out his consumption.
-The individual might be expected to maintain a stable or slightly increasing level of consumption.( he is an intertemporal optimiser).
The model suggests that in the early years of his life , he is a net borrower.
In the middle years of his life , he saves to repay his debts and to provide for retirement.
In the latter years of his life, he dissaves.
-The advantage of the Life-Cycle approach is that it takes account of the individual's accumulation of ASSETS. e.g. house, shares and adds it to their wealth. Income and value of assets can be liquidated if the individual wishes.
-However, the indicidual faces a budget constraint C<Y,
where lifetime consumption cannot exceed lifetime income.
We can employ the process of discounting with a certain rate of interest.
Discount rate = future income
1 + rate of interest.
Applying the discount rate over a stream of future period we get the PRESENT VALUE (PV) of future income.
C = K(PVt)
where i = individual
t = time
K = mpc
-The Life-Cycle approach makes the PV term operational by splitting it into components of income from labourand income from assets and property.
PVo = Y + Y
(1+r) (1+r)
where o = current period
If the capital mkts are efficient, assume that the PV of income from an asset is equal to the value of the asset itself, measured at the beg. of the period.
Y = a
where a = real household net worth at the beginning of the period.
We can further seperate out known current labour income from unknown, or expected future income.
PV = Y + Y + a
(1+r)
We also need to determine how expected labour income Y ...... Y might be related to current observable variables.
Assume average expected (e) labour income in time o , Yo, such that
Yo = 1 Y
T-1 (1+r)
where T-1 is the average remaining life expectancy and the term 1/(T-1)
averages the PV of future labour income over (T-1) years.
Expected labour income can be written as
Y = (T-1) Yo
(1+r)
PVo = Yo + (T-1)o + a
-But how is expectations of future income bought about?
Expected future income rises if current income rises and the current income is viewed as permanent. e.g. if a tax cut ids viewed as permanent then it will raise consumption present and future. If temporary, then current income will not increase as much.
Yo = Yo >0.
PVo =[1 + (T-1)]Yo + a
-Explanation of empirical facts from cross-sectional data:
1.Low income gps had higher apc than high income gps.
Low income gps high roportion of young and old hsehlds with high propensity to consume. High income groups have high proportion of middle aged hsehlds with a low apc.
2.Variations in apc when using time series data can be explained as follows. Any transitory income or boom is spread over indiv. remaining lifetime.
Consider:
Person with 20 more yrs to live receives windfall of £1000. Their revised permanent income increase by £50 p.a. Consumption in the yr the windfall is recieved increases by small amount. apc would then fall with higher income because consumption would have risen by a small amount, based on permanent income, but measured income has risen by £1000.
An unexpected fall in income in recession would be spread over indiv. lifetime, with borrowing and/or running down of past savings leading to a small cut in that indiv.s current consumption, causing apc to rise.
3. Therefore, the LCA like the permanent income approach has reconciled the flatter SR consumption function with the steeper LR C function(with constant apc).
Both theories imply that the mpc which is the slope of the consumption function, is lower in the SR than it is in the LR. IN Friedman's case any unexpectedincrease in income is not consumed but largely saved, whereas in the LCA it is spread over the consumers lifetime.
(thus if the life cycle approach is true then cross sectional studies will show that apc falls as income rises, which also explains the cross sectional budget studies showing MPC < APC.
Lifc Cycle approach , the elderly and bequests.
MODIGLIANI -
people save for retirement and draw down their savings during that period.
KOTLIKOFF AND SUMMERS-
Savings to provide bequests rather than provide for consumption when old. Amount of wealth in the economy far too large for people to be saving only for their retirement.
CONCLUDE: people saving mainly to pass wealth onto descendents.
DANZIGER, GAAG AND TAUSSIG-
elderly save more than the young. Inconsistent with life cycle approach.
But in Japan, elderly pool their wealth with their children. Drawing down wealth in thier retirement, but wealth not distinguished from that of their children who are saving.
WEALTH belongs to the C function.
-Drawbacks of the Life-Cycle approach:
1. Liquidity constraint
Every individual can borrow againt its future income but cant borrow from capital mkt. In boom - more likely to be employed and less constrained - makes SR less steeper. Can't borrow during recession.
2. Imperfect capital mkts
Savings earn interest but interest rate not the same for everyone.
3.uncertain of lenghth of lifetimes
sometimes want to leave bequests to heirs. Not plan to consume all resources over the lifetime.
4.Must take a/c of composition of the family over time.
consumption provided for the children before they go to work.
5.uncertainty about future income.
Not know lifetime labour income, and consumption to be made on the basis of predictions about future income. But note: expected lifetime labour income related to current disposable labour income . Can get around uncertainty by expected utility approach. Indiv. subjective probability about diffn. states. Pure uncertainty - do nothing. Risk uncertainty - measurable.
But some view this as not a good approach bec. people will arrive at certain bad events if it means not consume / behaviour optimality. Therefore, min level of consumption may not optimise serious modification.
Friedman: people gear their consumption behavior to thier permanent or long term consumption opportunities, and not to their current level of income.
e.g. worker paid weekly on Friday. Does not consume all on that day.
-Indiv. prefer smooth C flow.
note:
1. C on any day of the week would be unrelated to income on that part. day but would be geared to average daily income.
2. Income for period longer than a day is relevant to the C decision .
3. Nothing special about the period of time that requires the indiv. to plan consumption. C is planned in rrelation to income over longer period.
-PIA argues that consumption is a proportional to permanent income.
C = cYP
and k = F(i,w,x)
where YP = permanent disposable income.
The proportion c is determined by interest rate(i) , ratio of non-human to human wealth (w) , and a catch all variable (x) including age, taste.
c will increase if w rises in total wealth holding. This also increases indiv. security, since non-human wealth, such as money and shares, is assumed more reliable than human wealth, such as expected future labour income.
-Defn of permanent income:
the steady rate of consumption a person could maintain for the rest of his life, given the present level of wealth and income earned now and in the future.
-Distiction between permanent & transitory income:
Permanent: LT view of one's earning ability based upon sustained events.
Transitory: random increase or decrease in permanent income.
Unsystematic.
In general: person not certain about what part of any change in income is likely to be maintained and is therefore permanent, and what is to be transitory.
-Friedman ignored income from wealth.
He argues that indiv. utility max. gives a relationship beteen indiv. C and PV.
C = f (PV ) f>0
Multiplying PV by a rate of return r gives us Friedman's permanent income.
Yp = r . PV
This is the permanent income from the consumers PV which includes his human capital - the PV of his future labour income stream.
Friedman argues that consumers want to smooth out actual income into flat C pattern.. This gives the level of permanent C, Cp that is proportional to Yp.
Cp = K Yp
He goes further to define total or observed income as 2 components:
permanent income that the indiv. has imputed to himself, plus a random transitory income component which can be (-), (+) or (0).
Y = Yp + Yt where t transitory.
Total permanent C and random transitory C :
C = C + C
To infer what part of an increase in income is P or T is resolved by assuming that permanent income is related to the behavoiur of current and past incomes.
e.g. estimate permanent income as being equal to last years income plus some fraction of the change in income from last yr to this yr.
YP = Y + 0 (Y - Y ) 0< <1
= Y + (1- )Y
where is the fraction
Y is last yrs income.
note: 1. If Y=Y , then P income is equal to the income earned this yr and last year. Guarantees an indiv. who had always earned the same income would expect to earn that income in future.
2. If income rises this yr c.f. to last yr, then permanent income rises by less thancurrent income. Idiv. not know if rise in income this yr is permanent . Indiv. not immed. increase expected or permanent income meas. by the full amount of the actual or current increase in income.
-Friedman's approach makes the following assumptions:
1. No correlation between permanent and transitory income.
2. No relationship btwn permanent and transitory income.
3.No relation between transitory income and transitory consumption. Difference between consumption and consumer expenditure.
(detailed explanation later).
-Expanation of the X-section result that MPC < APC (even when the basis of the theory is that ratio of permanent C to P income is constant).
ref: diagram below
-As economy cycles above its trend growth path, the apc will move above and below the LR income line.
-Pt A: In an average yr when transitory income is zero then intersection of consumption with permanent income at Pt A.
Here, the economy grows steadily with no fluctuations, constant proportion of Yp consumed.
-Pt B: The point of lower income coincides with either recession or c.f. across hsehlds will refer to low income hsehlds. (lower than avaerage).
When we take a horizontal line onto the LR we obtain the permanent level of income Yp , but if a line is dropped vertically from B we get an income level of Y . (gives us the SR function above).
Y refers to the measured income of the individual which is less than the permanent level.
i.e. Y < Y
Why is this ?
1. hsehlds cross-sectionals taken into a/c.
At point B, indiv. has a bad yr . more income < permanent income component. Negative transitory component. But a fall in actual income should not affect people's consumption patterns.
2.Consumption expenditure may take place bec. of transitory income but wont be consumed in one period hence C at its permanent level. i.e Y .
3.Pt C: converse of Pt B.
Hsehlds with good years. ie. positive transitory income.
Business cycle reflects a boom period where actual income Y is grater than the LR average. (when have growth, the SRC function moves up).
But the permanent income is less affected by the rise in income and only rises to . Only if meas. income remained at for several yrs would permanent income be revised upwards to and consumption to . If this is not the case, then the proportion of income consumed will only be , i.e. the apc will have fallen below the initial level c set during the boom..
-Friedman's model xplains the X section budget studies and SR cyclical observations that indicate MPC <APC, as well as the LR observation, which is much steeper, that C/Y ratio is fairly constant. i.e. APC=MPC.
(SR function flatter and variable apc).
Little of upward revision plans when transitory income. Recession will cause apc to rise unless long -lived.
Advantage -
1. a/c random shocks to income transitory approach.
2.Explicit in the theory. LR permanent view of income.
Capital derived by utility max. as in LCA where Y which has to be discounted at current and future income.
-But the model is less satisfactory than LCA bec assets only implicitly taken a/c as determinants of permanent income and relies on less observable aspects of permanent and transitory income.
-Propensity to consume depends on relative income through time and between gps at one point in time.
-From time series data, theory argues that determinant of consumption is indiv. current income relative to his previuos peak income.
Consumers get use to certain level of C and if income falls, they resist the fall in C that is implied. i.e. apc rises in slump.
-Basic LR X function proportional to income, it still varies with the business cycle.
If income at Yo, higher than previous peak period, then consumption at E with C = kYo.
If next period, income falls to Y , C contracts not along C=kY , but by less to E . apc at Y > Yo bec. people reducing the proportion of income saved in order to maintain C at lower income.
-In a slump apc rises above k.
-In recovery, as income increases, C expands along SR C function E -E.
i.e. recovery in apc.
When the previuos peak income Yo is reached and surpassed, C expands along along C function. If , when income reaches Y , there is another recession and income falls, C contracts by prop less and apc rises.
-RIH explains empirical fact that if X sectional datataken, apc falls as income rises. Suggests low income gps aspire to average C patterns. The prop of income C or S depends on indiv. income relative to average income.
If indiv. has below average income, then his attempt results in C of higher than avearge proportion of his income. and vice versa. This will give a flatter curve observed for C functions using X section data.
*************************************************
SIMILARITIES AND DIFFERENCES OF THE CONSUMPTION THEORIES
- Consumers plan C expenditure on basis of lifetime income expectations, not on the basis of income received in the current peiod.
-If people can borrow and lend freely, C in any one time period should not be tied closely to income received in that time period. Desire to even out C in face of fluctuating income.
-ultimate constraint is wealth avail. to indiv.- in form of human and non-human assets.
-Explains consumption rather than consumer expenditure. Consumption includes the use of non-duable goods and services (as measured by depreciation). A windfall income even if spent on consumer durables is classified as saved implying a low mpc out of a change in current income.
-multiplier predicted will be small in SR bec. mpc in Sr. Govt policies in G and T ineffective esp if temporary.
-mpc lower in SR than LR. In PIH unxpected increase in income not consumed, but in LCH spread over consumers lifetime.
DIFFERENCES
---------------
1.KEYNES V. LCH PIH
Current disposable income main factor influencing consumer expenditure.
But 2 reasons why more important:
1.not all consumers able to borrow and lend in diffn periods to even out consumption.
2.estimates of future disposable income on which permanent income based, highly uncertain.
2.By product of Friedman's wk is attack on Keynesian SR dd management.
