Consumption Flashcards
We will look at 5 approaches of theories of consumption
Keynes
Modigliani - Life cycle hypothesis
Friedman - Permanent income hypothesis
Hall - random walk hypothesis
Laibson - inconsistent preferences
Approach 1: Keynes: 3 main characteristics
MPC between 0-1
Average propensity to consume (C/Y) falls as income rises; rich save proportionately more than the poor, saving is a luxury.
Income is the main determinant of consumption, not interest rates.
Keynesian consumption function
C = Cbar +cY
Or Cbar = autonomous consumption (when income=0)
c = MPC (used to be b)
Y = disposable income. (Used to be Y-T)
Keynes consumption function (KCF) in diagram form - that conforms to all 3 characteristics
Empirical validity of Keynes model - which event supports this
The Great Depression supports Keynes model, of income being the main determinant of consumption (Not interest rates)
Empirical application evaluation (2)
What is the main conclusion from this, and what is this idea known as?
- Secular stagnation during WW2.
Recall KCF’s 2nd characteristic suggests rich save more than poor (APC falls) (saving a luxury).
However in reality APC was more stable than KCF predicted.
- Kuznets found ratio of consumption to income was stable despite large increases in income (KCF says consumption should fall as income rises)
In real life, APC falls only in short run/cross section (explained by transitory incomes-Friedman). But in long term, ACF is stable. This is called the consumption puzzle.
How to draw this consumption puzzle
Short run consumption function is flatter as we have falling APC - consumption increases less as income increase
Long run consumption constant APC
Modigliani and Friedman’s main idea
People prefer smooth consumption, so don’t only account for current income, but expected future income.
Approach 2: Modigliani
Consumption depends on lifetime income, which varies systematically. (Systematically is a criticism - too simplistic)
Saving allows income to be transferred from periods in which it is high (young-working) to periods when low (old) , leading to the life cycle hypothesis
Life cycle hypothesis:
Consider a consumer who expects to live for another T years, has wealth of W (stock) and expects to earn income Y (flow) over the next R years until retirement…
Express this…
C = (W+RY)/T (Wealth x income x years worked) then divide by T to make smooth.
rewritten as
C= (1/T)W + (R/T)Y
which we can aggregate up further…
If we aggregate up, what is the equation and what do we get?
We get 2 MPCs
C= aW + βY
a is MPC out of wealth (stock)
b is MPC out of income (flow)
Once we have the aggregate consumption formula,
What is the key finding we can now make?
We can find APC.
C=aW + βY
APC = C/Y
C/Y= a(W/Y) + β
SR = wealth (stock) is less variable than income (flow); so high income (Y) means low APC. (flows are easier more flexible)
LR = wealth and income vary by similar proportion (higher correlated), resulting in stable W/Y and a constant APC.
(As the consumption puzzle shows!)
Life cycle consumption function graph
and how does it differ from Keynesian consumption function
In KCF, intercept was Cbar (autonomous consumption when Y=0)
Now, the intercept is drawn for a given with wealth, so intercept is aW.
Increase in wealth diagram, and when can this happen?
An increase in wealth would shift upwards, since aW is the intercept.
This can happen from the short run to long run (since wealth is constant in SR, vs variable in long run)
C,Y,W Life cycle diagram:
axis’, and what happens (explain 4 variables)
Y axis is £
X axis age
Consumption - constant, so vertical line as smooth.
Wealth - increases while working since saving, until retirement it falls. (Dis-saving)
Income - constant, so vertical line, until retirement when immediately drops to 0. (PIH later questions smooth income…)
Saving - represented by the box between income and consumption. (Since what is not spent is saved)
Approach 3: Friedman - complements Modigliani (both argue consumption does not depend on current income alone, future expected income)
What was Friedmans main theory?
Permanent income hypothesis - consumption should depend on permanent income Yp
How does Friedman’s Permeanent income hypothesis differ from Modigliani?
PIH allows for non-regular income patterns over a lifetime.
I.e people can experience random and temporary changes in income.
(Modigliani’s is oversimplified in assuming income is smooth, as we see in slide 11)
Friedman looks at the 2 components of current income.
Current income formula
Y = Yp +Yt
Yp is permanent income.
Yt is transitory income
Permanent income vs transitory income
Permanent income is income that people expect to persist into the future, so essentially the average income.
Yt is transitory income, people do not expect to persist i.e fluctuations around the average, e.g an especially good harvest for a farmer so gains higher income.
Friedman’s intuition
Consumption should depend primarily upon permanent income; saving or borrowing can be used to smooth out transitory increases or decreases in income to keep consumption smooth.
e.g you would respond differently to a permanent increase in income of £10,000 per year, compared to a transitory one-off increase of £10,000.
So accordingly to this, what does the consumption function become
C= aYp
a is the fraction of permanent income consumed.
(I,e MPC for permanent income)
PIH also rejects the KCF. Why?
PIH suggests KCF uses the wrong variable in APC.
KCF APC = C/Y = Cbar/Y + c
(Autonomous consumption/Y + MPC)
PIH APC
APC = C/Y = aYp/Y
So under PIH conditions.
What happens to APC
if Y=Yp ,Y>Yp or Y<Yp
If Y=Yp, a=APC
if Y>Yp - current income>permanent income, and so APC falls below a temporarily (Y=Yp+Yt, and so income is temporarily higher as a result of Yt, reducing consumption, saving luxury!!!)
if Y<Yp - current income<permanent income, (so must’ve been a negative Yt, thus reducing current income (Y), so APC rises above a temporarily
Explaining cross-sectional studies (Why APC varies across HHs)
If all variation in current income (Y) was purely due to changes in Yp permanent income, all would have the same APC (C/Y).
But it doesn’t. Current income includes Yt, and different HH will have different Yt, thus explaining why richer HH have a lower APC.
Explaining time series studies (APC in short run vs long run)
In the short run we have fluctuations in income to reflect transitory changes (Yt) , so APC may be higher or lower depending on whether Y>Y or not. (Slide 13 e.g APC higher if Y>Yp)
In long run, fluctuations in income are primarily due to permanent income. (constant APC)
Approach 4: Robert Hall
Robert Hall uses this PIH (consumption based on current and expected future income [permanent income]) but adds rational expectations.
What are rational expectations
People use ALL available info to make optimal forecasts about future.
(Different from adaptive expectations eq 4 - which is naively applied, Etπt+₁=πt simplistic!!!)
Hall’s Random walk hypothesis
Consumers face fluctuating income and seek to smooth consumption overtime, during this they receive news and surprises which makes them revise their expectations.
These surprises are random, and so changes in consumption are unpredictble too, so consumption follows a random walk.
What do consumers base their level of consumption depend on?
Expected lifetime income.
How can their level of consumption change (2)
- News
- Surprises
Unpredictable thus we say consumption follows a random walk.
Implications for policymakers for Hall’s approach.
ONLY UNEXPECTED policy changes influence consumption , since Hall’s approach uses PIH and rational expectations which accounts for EXPECTED events already.
(So it is saying hexpected policy change should have no change on consumption)
Example of policy being ineffective in influencing consumption since already accounted for in their PIH & rational expectations.
Economic stimulus act 2008. In response to the financial crisis, provided 115bn tax rebate (tax refund i.e cut) in attempt to boost consumption.
What did consumers do instead?
HH chose to use money to pay off debt rather than consumption.
What did Taylor suggest shouldve been done instead?
Extend the tax rebate (to increase permanent income)as opposed to a one-off payment (this is transitory income) in order to get the desired effect (increased consumption)
Parker’s contested interpretation of the Economic Stimulus Act overall
Since the rebate was one-off and predictable, it should’ve made no difference at all to consumption. (Since accounted for in R.E…)
However Parker found HH still spent 20/30% of the rebate on consumption, thus rejecting PIH & rational expectations.
Approach 5: David Laibson (hint. Behavioural)
So far we assume consumers are rational.
Here, consumers doubt their decision making capabilities, and have INCONSISTENT PREFERENCES.
What is the idea of inconsistent preferences?
Consumers may alter their choices overtime.
E.g consumer planning to save for retirement, but then allows herself one last spending splurge.
Consider an increase in elderly population -
What does the life cycle model predict will happen?
- What further questions will be raised
Savings will fall, as more old people, so spending and running down their savings.
2.
Who will buy their assets? Since more old now. May habve to accept a lower price, and thus compromise on their expected standard of living in retirement.
This also arises the question to how the state should address this. E.g more generous pensions?
Will youth have to pay higher tax to contribute to this?
