EFB338 Contemporary Application of Economic Theory
EFB338 Contemporary Application of Economic Theory
EFB338 Contemporary Application of Economic Theory
Semester 2, 2010
Matthew Robinson
Page
Matthew Robinson 1
.9.1.Types of labour market policies.........................................................................................................................45
.9.2.Types of unemployment and job search.............................................................................................................45
.9.3.Active Labour Market Policies (ALMP).............................................................................................................46
.9.4.Australia’ work for the dole...............................................................................................................................47
.9.5.Other empirical findings....................................................................................................................................48
ECONOMIC GROWTH (JAY)......................................................................................................................................49
10. INTRODUCTION.............................................................................................................................................................49
11. NEOCLASSICAL / SOLOW MODEL.....................................................................................................................................49
.11.1.Solow model.....................................................................................................................................................49
.11.2.Mankiw, Romer and Weil (MRW) & criticism.................................................................................................49
.11.3.Overview of weakness......................................................................................................................................50
.11.4.Improvements to the model...............................................................................................................................50
12. EASTERLY & LEVINE’S 5 STYLISED FACTS.......................................................................................................................52
.12.1.TFP is responsible for growth..........................................................................................................................52
.12.2.Divergence in per capital income over time....................................................................................................52
.12.3.Factor accumulation is persistent, but growth is not.......................................................................................52
.12.4.Economic activity is highly concentrated.........................................................................................................53
.12.5.National policies care closely associated with long term economic growth...................................................53
EXAM PREPARATION..................................................................................................................................................54
13. SAMPLE QUESTIONS FROM BLACKBOARD...........................................................................................................................54
.13.1.PART ONE.......................................................................................................................................................54
.13.2.PART TWO.......................................................................................................................................................54
.13.3.PART THREE...................................................................................................................................................54
.13.4.PART FOUR.....................................................................................................................................................54
MY ESSAY........................................................................................................................................................................55
Matthew Robinson 2
The Shadow Economy (Benno Torgler)
1. Tax compliance
.1.1. Introduction
()a Importance of taxes
Taxes are considered to be what we pay for a civilised society.
Taxes are key to the operation of a civilisation.
Although it is supposed that noone likes paying tax, without it, there would be problems in
public finance, law enforcement, organisational design, labour supply, etc.
• reduces tax collections, affecting the amount of tax compliant taxpayers receiver or
the services that citizens receiver
Matthew Robinson 3
Figure 1 - Severity of punishment as a function of the seriousness of the offence
In Figure 2 above, committing the offence is not worthwhile, even for an amoral person.
However, where the expected return is positive (ie where the E(Punishment) is below the 45-
degree line), the benefit of committing the crime exceeds the expected cost.
A rational person will maximise the seriousness of their offence such that they will commit the
offence with a seriousness where marginal benefit = marginal expected costs (ie until they
achieve X* in Figure 3).
Matthew Robinson 4
Figure 3 - When does crime pay?
• the criminal is informed – he or she knoes the costs, benefits and probabilities
associated with crimes – transaction costs are not involved with obtaining this
information
Matthew Robinson 5
Many crimes occur where the person has diminished rationality.
Two terms are relevant:
Matthew Robinson 6
• the tax rate on reported income: t
• the penalty rate on each dollar that the person was supposed to pay, but did not: f.
If the taxpayer is caught, their payoff is:
Payoff = Y – t D Y – f ( t (Y - DY))
Payoff = Income – tax paid – penalty on tax not paid
If the taxpayer gets away with evasion, their payoff is:
Payoff = Y – t D Y
As such, the expected value of the taxpayer’s evasion can be represented as:
E(V) = (1 - p) (y – t DY) + p(Y – t DY – f (t (Y – DY)
E(V) = (probability of no audit x payoff without audit) + (prob of audit + caught payoff)
An individual will seek to maximise their payoff. Because the only equation that is up to the
individual is the amount to declare, we maximise for the DY. The maximised equation is:
pxf=t
In other words, the taxpayer will evade tax until the expected penalty equals the tax.
Empirical evidence suggests that people actually report more income than this model predicts.
This implies that there is a high degree of voluntary compliance.
• YD is declared income
Matthew Robinson 7
E(V) = (1 - p) (y – t DY) + p(Y – t DY – f (t (Y – DY) – d (Y – DY)
where d is the disutility from evading tax, which is greater as the amount increases.
When maximised, the formula is:
pf=t-d
• audit selection is endogenous – the integration of the audit selection rule to include
factors such as previous evasion is likely to lead to better results
• evasion is considered to be a one period game – litratutre suggests that where there
are voluntary public good contributions, contributions decline with each repition
• it does not take into account social interactions (conditional cooperation) – ie if the
government can affect the social norm of compliance through:
• there are numerous extensions that complicate the analysis and make clear-cut
results hard to isolate
• Bordignon (1993), who assumes that taxpayers compute whether there is a fair
trade-off between the diminished private consumption and the government provision
of public goods; if the terms differ from fair terms, the taxpayer will evade tax to even
out the unfairness, subject to the risk of getting caught.
• Erard and Feinstein (1994) incorporate shame and guilt directly into the taxpayer’s
utility; they assert that guilt is felt for underreporting even if not caught and ashamed
if caught; these are hard to work with as they are unobservable. There is the
possibility of using nueroscientific tools to explore guild and compliance.
Matthew Robinson 8
.1.4. How to measure non-compliance (methodologies)
Because of its nature, tax compliance is concealed, meaning that economists must be creative
when attempting to measure compliance. As such, empirical evidence is relatively rare.
• 40% underpaid their taxes – about 25% underpaid by more than $1500
• 7% overpaid
• the information is purely based on the taxpayer’s report and the IRS
examiner’s detection of evasion, which may not discover all evasion
• the difference between official labour force pariticpation rates and estimates of true
participation rates
• look for traces of unreported income in monetary aggregates (ie the use of cash in
the shadow economy.
The IRS estimates there was a tax gap of between $257 to $298 billion in 2001, indicating a
non-compliance rate between 15.5 to 16.6%.
The tax gap in switzerland is estimated to be 23.5% of all income tax for the years 1970 to
Matthew Robinson 9
1995.
Problems include:
• driven by assumptions.
()c Surveys
(c.i) Introduction
Surveys are designed to elicit taxpayer’s attitudes about their reporting (tax morale).
These can be adapted to estimate non-compliance.
Surveys are amenable to collecting socio-economic and demographical information,
which helps to examine and test theories on tax evasion.
Surveys can be helpful to compare countries, or compare changes over time.
The ability to access more sophisticated statistical techniques have made surveys a
more popular mechanism to collect relevant information.
Problems with surveys include:
• “Does your employer deduct your personal income tax from your monthly
salary?” An answer of no may indicate tax evasion.
• “Please indicate who pays your tax on personal income or your tax on small
business.“ An answer of nobody may indicate tax evasion.
(c.iii) Examples
Matthew Robinson 10
The Taxpayer Opinion Survey provides a broad set of taxpayers’ opinions on aspects
of the tax system, including the IRS, tax evasion and cheating on taxes.
Questions include “Within the past five years or so, do you think you maight have
overstated any deductions or expenses … even by a small amount?” and “Within the
past five years or so, do you think you might have left some reportable income off
your federal tax return – even just a minor amount?”, with answers being on a scale
of 1 to 4 being “definitely have not” to “definitely have” respectively.
(iii.B) Afrobarometer
Worded the question as: “We would like to remind you that your responses to this
interview are confidential. Here is a list of actions ordinary people are taking in a
political system. For each of these, please tell me whether you have engaged in this
activity or not?”
For “Avoid paying income taxes”, the variables were coded as follows: Yes, often;
Yes, a few times; Yes, once or twice: (0), No, but would do it if had the chance: (1),
No, would never do this: (2)
(iii.C) Latinobarometro
“Could you tell me if recently you have known someone or have heard someone you
know comment about somebody who has: Managed to avoid paying all his tax”
(1=yes, 0=no).
“On a scale of 1 to 10, where 1 means not at all justifiable and 10 means totally
justifiable, how justifiable do you believe it is to: Manage to avoid paying all his tax?”
Matthew Robinson 11
NB: -ve correlation means less compliance
• control focus variables – by controlling all other variables, you can test the
effect of a particular variable
• external validity – the result from the lab experiment may not generalise well
to the larger, ongoing outside the lab. Thus, insights may not be able to be
extrapolated to the world beyond. The choices made in the experiment may
be somewhat artificial – therefore the external validity of the experiment
may be flawed
• Subjects know they are being watched, and may alter their behaviour
• The provision of financial incentives (and its quantum) may affect behaviour
• Often students are used for these experiments; problems could include a
lack of experience and the possibility for a higher IQ than the population
generally or may come from families with a higher income; age ranges may
also be limited. (NB: evidence to suggest that students’ responses are not
different from those of other subjects – indicating cognitive processes of
students are not materially different from those of other people). A reason
Matthew Robinson 12
for using university students is purportedly their ability to easy understand
abstract problems.
• subjects receive income each round and pay taxes on the reported income
(if the experiement has a public good, all taxes finance the good)
(ii.A) Introduction
Field experiments allow researched to evoke real processes and behaviour, which
allows a better test of the effects of variables in the real world. This allows to
formulate practical policy responses.
An example of a field experiment would be for a tax authority to send out a letter
about the importance of tax (or provide awards for compliance) to 50% of the
population, while leaving the remaining 50% as a control, thereby testing the effect of
the action on compliance.
The benefits from field studies include:
Allows social and economic interaction
No observable experimental effect as participants are unaware of having
taken part in the experiment
More representative than laboratory experiments
Matthew Robinson 13
Allows researchers to test specific policy alternatives, such as the effects of
a higher perceived audity probability or the effects of moral suasion on tax
compliance
The problems with field studies include:
Less control – may make causality harder to determine
Need the cooperation of tax authorities
High transaction costs to deal with tax authority
Sensitivity of the tax filing data
Consumes a great deal of resources
Sometimes difficult to develop a treatment that can be hard to get approved
by the tax authority – thus unorthodox treatments cannot be used.
Time can be a massive problem – the experiment has to be prepared before
participants receive their tax forms, and it can be over a year until all tax
forms are returned.
Field experiments are generally only used once – it’s hard to know the long
term effects because there are not inter-temporal studies.
()d Income
Higher income leads to higher reported income (but the lecture slides are unclear as to
whether this amounts to compliance).
Matthew Robinson 14
()f Tax complexity and uncertainty
Greater taxpayer uncertainty about true taxable income leads the taxpayer to report a higher
taxable income
Greater uncertainty about the tax, penalty and audit rate raises compliance.
This may be because there is a greater perceived punishment.
()h Institutions
How tax revenues are used and the decision process (ie is it voted on, as in Switzerland)
affects compliance.
Poor institutions may to less compliance.
Trust in the state leads to greater compliance.
Respectful treatment by tax authorities leads to greater compliance. Treating tax authorities as
service departments leads to greater tax compliance.
Direct democracy and local autonomy are key factors – they set the rules between the state
and taxpayers. More direct democracy and stronger local autonomy lead to significantly
higher tax morale; it is posited that this is due to the fact that there is strengthened acceptance
of government decisions and hence the willingness of the public to contribute towards it, as
well as making taxpayers feel responsible for the decisions.
Fairness and transparency increase taxpayer’s propensity to contribute to public goods even if
they do not benefit from them.
If people velieve that others are honest in paying taxes, this will increase their willingness to
pay tax – this brings in the notion of reciprocity. It is posited that this represents the
diminished social disapprobation associated with evasion.
.1.6. Conclusion
The plethora of factors that influence compliance makes a unified theory of tax compliance
difficult to achieve.
Matthew Robinson 15
Matthew Robinson 16
2. Corruption
.2.1. Introduction
Corruption is defined as an illegal payment to a public agent with the goal of obtaining a benefit or
avoiding a cost.
Because the state controls the distribution of valuable benefits and imposition of costs, there is an
incentive for people to influence those who have discretionary power. The higher the economic
rents associated with the benefits received / costs avoided, the more incentive a person has to
evade regulations, and the higher the payments can be.
Corruption can undermine the legitimacy and effectiveness of government.
Inherent in this is the principle-agent problem.
Matthew Robinson 17
• Freedom of association and of the press, allows exposure of abuses
• Religion may influence how individuals view loyalty to family v society generally
• The church may play a role in monitoring and denouncing abuses by state officials.
.2.5. Measurement
()a General methods
• published sources – newspapers and internet
Matthew Robinson 18
• Voice and accountability – measures the political process, civil liberties and
political rights
• Rule of law
• International Country Risk Guide – yearly data on corruption – assesses the corruption
within the political sustem.
The first two indices are highly correlated (90%).
Potential problems / biases:
• cultural background
• does the survey consist of locals or expats – more objective or ignorant of culture?
• do respondents consider the problem of corruption relative to other problems within the
country?
• Hundriks (1999) used anecdotal evidence to find that 94% of Tiwanese tax administrators
had been bribed (76% in India).
• Tracking entertainment and travel costs in China – although does include corrupt
payments, it also includes legitimate expenses
• measuring the amount of rice distributed b the amount actually received by households in
an Indonesian anti-poverty program
Matthew Robinson 19
Proposed to work against:
• extortive corruption – government official had discretion vis-à-vis a service / permit (ie in
Phillipoines a tax inspector would assess an unrealistically high payment on the taxpayer;
it was hard to determine what the actual payment due was, and hard to appeal. Bribe of
official to give correct figure.) A free press may identify the corruption and raise the costs
for the bureaucrat – it has a platform of voicing complaints. However, an argument that
media is only concerned about large events.
()a Model
Dependent variable is corruption measured by the International Country Risk Guide
Main independent variable is a measure of press freedom (with 0 being the best and 100
worst).
Controls:
• GDP
()b Results
The study found that greater freedom of press was correlated with less corruption (negative
coefficient). Results were statistically significant.
The study controlled for OECD / non-OECD differences and the effect was present in both cases.
Multiple indices of corruption and press freedom were used to ensure robustness – the effect still
existed.
They used press freedom at the start of a 5 year period, and the other variables over the period, to
try to establish causality.
Matthew Robinson 20
.2.7. Case Study – Gender and Public Attitudes towards
Corruption and Tax Evasion
()a Aim
To test whether women are more averse to engaging in corruption and tax evasion compared to
men.
An analysis necessarily involves controlling for the opportunity to evade taxes. If there were
identical opportunities, the null is that women would be just as corrupt as men.
Why are women proposed to be less corrupt:
• Alcohol and drug abuse are more common among men than women.
• Studies have found that women are generally more likely to contribute to
public goods or common objectives (with some caveats).
()b Method
Used individual-level data from the World Values Survey and the European Values Survey for 8
western European countries from 1980 to 2000.
Used questions relating to the acceptability of corruption and tax evasion. Justifyability of tax
evasion is highly correlated with the Transparency International Corruption Perception Index
(0.358 – statistically significant) and Quality of Government rating (0.38 – statistically significant).
If the participation hypothesis is correct, then women should become more corrupt as they attain
greater equality.
Using countries in the same region should reduce cross-cultural comparisons.
()c Results
Women are less likely to engage in corruption and tax evasion, even after controlling for increased
opportunities to become involved in such activities. Results show very little support for the
opportunity theory.
Gender differences hold for each period and each country, despite the different opportunities for
illicit activities for men and women at different times and in different countries.
The justifiability of tax evasion indicates an increase of the marginal effect of gender from 4.2 to
7.6%.
Matthew Robinson 21
.2.8. Case Study - Winning isn’t everything: Corruption in Sumo
Wrestling
()a Introduction
In sumo wrestling competitions, there are 15 rounds, and winning 8 rounds guarantees that you
move up in the rankings, while only winning 7 guarantees you will move down in the rankings.
The average increase in rank for a win is about 3 positions, however there is a sharp non-linearity
for a win that takes you from 7 to 8 – a movement of about 11 positions.Thus, there is a much
larger incentive for a person to win if they have a 7-7 record compared to a 8-6 record.
Moving one spot in the rankings is usually worth $3000 a year.
The study examined the behaviour of participants.
()b Results
The study found that the distribution of wins in each tornament was not normally distributed – there
was a disproportionate quality of wrestlers who won exactly 8 matches (26%). This compared to
12.2% for exactly 7 wins. Under a normal distribution, these would be equal.
The authors also conducted a multivariate analysis using a probability model. The dependant
variable was winning, and the independent variables “bubble” which was 1 if the wrestler was on
the margin, -1 if the opponent was on the margin, and 0 if both or neither were on the margin.
They also controlled for ranking differentials.
They found that for the last round, wrestlers on the threshold were 25% more likely to win than
expected. The effect was present in rounds 12 – 14, but to a lesser extent, and were non-existent
in round 11.
It is arguable that the effect can be attributed to increased effort.
However, in two periods media attention focussed on match fixing; in those situations the effect
completely disappeared. This can be attributed to the deterrence model, such that an increased
probability of getting caught will make it unprofitable to cheap.
If there is collusion, corruption may be positively related to the frequency in which the wrestlers
expect to meet in the future – if they meet in the future, there is a chance for punishment for
breaking an agreement.
Top ranking wrestlers may be less likely to be willing to throw matches because there is a large
tournament prize. The study found that if the person is in contention for winning the tournament,
the bubble benefit disappears.
There were very high win rates where wrestlers were in the same ‘stable’ of wrestlers, indicating
that stables may collude between themselves to optimise utility.
The study also found that the first meeting between the 2 wrestlers after a win for the person on
the bubble, the person who was on the bubble was 7% less likely to win. This indicated that they
would allow the player to win next time (alternatively there could be monetary compensation).
Industry insiders made public the names of people who were corrupt and clean. When comparing
matches factoring in these results:
• Where the person on the bubble was clean, against a corrupt wrestler,
there was a negative, statistically significant relationship, indicating that they lost more.
• Where it was clean vs clean, there was a positive, but not statistically
significant benefit to being on the bubble – this may indicate effort.
Matthew Robinson 22
Incentives are important.
It is suggested that non-linearity in payoffs allow corruption to be more prevalent.
Increased media scrutiny helps to deal with corruption as it increases the expected punishment
and probability of detection.
Matthew Robinson 23
3. Money Laundering
.3.1. What is money laundering?
‘Money Laundering’ is derived from the Mafia’s attempt to ‘launder’ illegal money via cash intensive
washing salons in the 1930s.
It is defined as the process of transferring money between the official and the unofficial economy.
The IMG estimates that 2-5% of the world’s GDP stems from illicit sources and therefore must be
laundered.
Money laundering is estimated to be worth approximately AUD2.8 billion. This is largely attributed
to fraud / tax evasion, and then illicit drugs. This is typically invested in:
• Professional services 7%
• Gambling - 16%
Matthew Robinson 24
• Executive opinion survey that avoids the limitations of official statistics
Matthew Robinson 25
• Correspondent banking: sender – bank of sender – SWIFT – Bank of
receiver – Receiver
• Gold & diamond market – can be purchased easily, with a high intrinsic
value, and relatively compact.
Matthew Robinson 26
.3.5. Theoretical model
()a For launderer
Based on conventional criminal theoretical approaches, the expected utility of the activity can be
written as:
E=U[(1-p)(R-C) + p(-C-F)]
= probability of getting away * (expected return [ R=(1+r)Y ] – cost of money laundering [C=cY] +
probability of getting caught * (cost of money laundering and punishment)
There will be an optimal point, which is:
Y* = [m (1 - p) - c] / [2 p f]
This will mean that an increase in the punishment, fine rate or cost will reduce money laundering.
• Money laundering merely a symptom of illegal activity – so addressing the illegal activity
may be more effective.
Matthew Robinson 27
Matthew Robinson 28
Business cycles and monetary policy (Stan Hurn)
4. Business cycles
.4.1. What is a business cycle?
3 key elements:
.4.2. Theory
()a Keynesian
Keynes emphasises an autonomous reduction in investment generated by a lack of confidence,
but did not set out how this would work.
A mechanism is one which consumption is smoothed and if investment is entirely driven by animal
spirits, then national accounts identity shows that output would be a highly persistent process with
animal spirits as the driving force.
Matthew Robinson 29
• An unexpected growth of the money supply and the certeris paribus higher nominal prices
and wages that go with it, make producers and wage earners overestimate the real prices
and wages for their products and services. These overestimated prices and wages induce
an increase in both production and the necessary labour supply, thereby creating an
economic boom.
• In the reverse situation, when money growth unexpectedly declines, the process too is
reverse: real prices and wages are underestimated because of the certis paribus lower
prices, and production and labour supply will be down. In this case there is a slump in
economic activity.
• Natural rate of employment / output assume A & K are fixed in the cobb-douglas
equation labour market equilibrium determines the output equilibrium.
• The effects on production and labour supply will be especially large if the perceived
changes in real prices and wages are regarded as temporary, as this will give rise to an
inter-temporal substitution of labour supply: in order to maximise leisure and income,
producers and labourers will work more when prices and wages are perceived to be
temporarily high and work less when they are perceived to be temporarily low. This
emphasises the importance of expectations in the formation in the building of models of
the macro-economy.
Problems with the basic Friedman-Lucas model:
• There will only be prolonged periods of above and below average growth under the
(unrealistic) assumption that people systematically err in overestimating prices in the
boom and underestimating them in the bust. Taking inflation into account, there will be no
periods of booms and recessions but only random deviations from trend growth.
• Since changes in the inflation rate are usually quite moderate, it is not realistic to consider
them to be the cause of major short term fluctuations in labour supply. Further, the ability
to adjust working hours may be difficult.
Matthew Robinson 30
The basic RBC model predicts that a positive temporary technology shock will result in rises above
their long-term trend in output, consumption, investment and labour. The rise can be persistent
because the additional investment means higher capital for future periods (this is referred to as
internal propagation).
Criticisms with RBC include:
• Does the economy experience large, exogenous productivity shocks in the short run? This
can be tested by plotting changes in output with the solow residual, which should
represent TFP.
• Are wages and prices fully flexible in the short run? ie are prices sticky – this is the basis
of New Keynesian theory.
• Menu costs – changing prices is expensive, so firms will not change them if the cost of
changing them is greater than the increased revenue. Sticky prices may be optimal for the
firm even though they are undesirable for the company. This assumes monopsonistic
competition, as the firm must have some control over its prices.
• Coordination failure: In recessions, output is low, workers are unemployed and factories
sit idle. Markets could clear if people were willing to cut their price or wages, but are
unwilling to without knowing that others will do the same.
• Staggered wages and prices: all wages and prices are not able to adjust simultaneously.
Pricing decisions cause the price/wage level to adjust slowly. Further, as individuals and
firms know wages / prices are slow for a while, so are unwilling to accept this.
Matthew Robinson 31
5. Monetary Theory
.5.1. Introduction
Models that allow output to be suboptimal (ie those that do not require the market to clear) allow
for economic policy to be used to improve welfare through the persistence of shocks and the
amplitude of the cycle.
Policy is always subject to lags, such as:
• inside lag – when it is recognised that policy should be applied to when it actually is
where the ay term is the persistence in output, the b & m terms are the effect of previousl and
current monetary policy and ε t is a shock that is autoregressive distributing model, that is
independently identically distributed (ie ah a mean of 0). Under such a model, we assume mt is set
prior to observing ε t, so the goal is to minimise the variance of yt. The optimal variance
minimising policy is:
b a
mt =− 1 mt−1 − 0 yt−1
b0 b0
If this is substituted back into the yt equation, yt equals the shock term, ε t. Such a policy removes
all persistence in output. This may be seen as the outcome under discretionary policy, sincnce it is
what would be chosen if there was re-optimisation each period.
()b Problem
However, if b1 > b0, we have an unstable difference equation, where mt will be increasing in
absolute value at an explosive rate. To deal with this, a ‘correction’ is applied to the previous
period’s policy such that:
yt = a0 yt-1 + b0 mt + (b1 + v) mt-1 + ε t
Matthew Robinson 32
Under this assumption of non-stationarity, which is in fact explosive, the growth rates of mt and
mt -mt-1, the variance of y is growing over time, such that the amplitude of the cycle is increasing
merely due to slightly mis-estimating b1.
• Even if the structure of the model is correctly perceived, computational difficulties will
defeat the majority of individuals. Further, it is assumed that rational individuals do not
make systematic errors, however in reality this may not be accurate. Also, it assumes the
people have sufficient information.
• the impossibility of discovering the correct model – there may be substantial disagreement
about the structure of the underlying model. The Lucas critique states that the behaviour
of rational individuals depends on the model adopted, and therefore changes in policy will
have unknown consequences because it will change people’s behaviour – ie encode
policy in behaviour.
• challenging the incentive to acquire information. If the price incorporates all possible
information about the thing, and therefore price = value, then there is no need to gather
information. Rational individuals will merely get an average of published inflation
predictions. The concern that there will be no incentive to gather information.
Under rational expectations theory, the output solution does not contain any monetary policy
parameters, indicating that monetary policy is irrelevant.
This conforms with the Sargent & Wallace (1976) proposition that output only deviates randomly
from its natural level. This is called the policy invariance proposition. Under this model (and
rational expectations):
• expected monetary policy cannot affect the level of economic activity in the model, since
output does not depend on any of the parameters of the monetary policy rule.
• Unexpected monetary policy does affect economic activity in a random way (because it is
random, a rule would reduce the variance of the residual).
• The use of ‘monetary surprises’ this leads on to the modern arguments for a rule.
Matthew Robinson 33
than would otherwise be the case (ie if policy-maker’s hands were tied).
The government has an incentive to create unexpected inflation to put output above the natural
rate. This creates a societal loss, because inflation is above the social optimal level. This is
attributed to Barro and Gordon.
Rational agents will expect this, and therefore the economy will have an inflation bias. This is
eliminated by the use of a rule because the government does not have the ability to affect policy.
Thus, provided that the rule is credible, a rule will be superior.
It is suggested by having an independent central bank, which is more adverse to inflation than the
rest of society, inflation will be lower when an independent central bank is delegated monetary
policy.
New Zealand attempts to induce policy makers to be more inflation adverse than society by
making their salaries tied to low inflation (often described as Walsh Contracts).
However, having a central bank be completely adverse to inflation would be detrimental if there are
disturbances to aggregates supply, which would result in an increase in the variance of output.
Central banks have an incentive to pay a game where they minimise inflation for multiple periods,
but then cash in their reputation to follow society’s preferences. They then re-build their credibility.
Matthew Robinson 34
6. Monetary practice
.6.1. Areas of agreement
Economists generally agree that:
• Monetary policy should be transparent (one way of doing this is to delegate the power to
an independent policy maker).
• An appropriate policy instrument is nominal interest rates. The way the economy adjusts
to a new equilibrium with constant inflation is encapsulated in an interest rate rule.
• A committee pools the disparate knowledge of its individual members such that it obtains
a wide view
• Committees are less likely to adopt extreme or idiosyncratic positions – they will tend to
inhabit the middle ground.
• information inclusive strategy – such that the central bank shouldn’t disregard information
so they have the flexibility to react to new information
• increased accountability.
Matthew Robinson 35
t=
i α
+β
ππ
( π
t−)+
* β
(
yyt−
y*
)
If output and inflation are at target levels, then the interest rate will be a. The real interest rate is a
minus the target level of inflation.
Taylor suggested that β π = 1.5 and
βy =0.5. These coefficients are much more contravertial
than the basic theory behind the taylor rule. However, there is agreement that β π> 1. This is
because you need to have sufficiently high interest rates that you are not seeing a decrease in the
real interest rate.
()a Inflation
There is debate over the index that should be used. Some argue that the headline figure of CPI
should be used as that reflects the price that needs to be paid by consumers, while others argue
that underlying CPI should be used because that strips out changes in prices of volatile items (the
argument is that there is no point using monetary policy to combat temporary price rises).
()b Target
A target of 2% is popular, but there is little to suggest that an inflation rate of 3% is much more
detrimental.
• they can undermine central bank credibility – if there is not confidence that the central
bank is not completely focusing on low inflation, then inflation expectations are not likely to
be anchored.
Matthew Robinson 36
• Monetary policy has limited influence on unemployment in the long run, so why make the
central bank responsible for this?
Arguments in favour of dual mandates:
• Central bankers do care about GDP and unemployment, so why hide it?
• There may not be any internal inconsistency with inflation targeting and a dual mandate.
Low inflation helps to boost economic growth over the long run by saving time and energy
associated with dealing with high inflation: having to reset prices regularly, re-write
contracts to deal with inflation, etc; facilitating long-run decision making, with consumers
and businesses not having to worry about uncertainty about the future price level;
enhancing the price signals and thus the functioning of the market.
• it provides an explicit metric that the public can use to assess the central bank’s
performance
• Which price index to use? The choice in indices can have large effects on the result.
Maybe it would be advisable to look at the trends in a number of indices.
• There is little indication that 3% inflation is significantly more harmful than 2% - the priority
should be to ensure stable growth within moderate levels of inflation.
Matthew Robinson 37
Policy maker may lean against an asset bubble by raising rates, effectively taking out insurance
against the risk of an asset bubble.
However, it is often argued that the cost of departing from optimal monetary policy is much more
costly than the benefits achieved from leaning against the bubble. This is especially so given that
bubbles have not proven to be particularly sensitive to the level of short-term interest rates.
Tools that are often used are:
• the bully pulpit: public question the assumptions underlying the bubble and threaten
policy to prevent the bubble.
• Monetary policy – tighten monetary policy to lean against the wind / bubble.
Matthew Robinson 38
Human capital (David Johnson)
7. Human Capital Accumulation
{Most of the content about private and public returns to schooling are included in the essay at the
end of the notes}
• identical twins
• class sizes can be reduced without disturbing the overall structure of the school system
• teachers are usually supportive (may have their own interests in mind).
Smaller class sizes are said to improve achievement because it allow more time to be devoted to
each student. Lazear (2001) provides a formal model which states that a person learns more in a
smaller class because each student is disruptive for a certain percentage of the time, which means
that the greater the number of students, the greater proportion of time they are going to be
disruptive.
It is hard to measure the effect of class size, because assignment is usually not random:
• If high achieving students are put into small, advanced classes, estimates will be upward
biased
• if low achieving students are allocated to small, remedial classes, estimates will be
Matthew Robinson 39
upward biased.
• allocation may be due to family attributes, such that wealthier families may move to
districts with smaller class sizes.
Empirical evidence:
• Hanushek (2003) argues that increases in school resources are an ineffective way to
improve school quality – finds improvement is not consistent.
• It is possible that sizes are close to optimal, such that further reductions have little
marginal effect, or are so small they are difficult to detect in small samples.
The positive effects (if any) of class sizes are thought to be outweighed by the large cost.
• students were not randomly assigned to schools, so there may still be school effects
Matthew Robinson 40
2 Contextual interactions: propensity of a student to behave in some way varies with
background characteristics of the group members 0 eg individual achievement varies with
the socio-economic composition of the group.
3 Correlated effects: students in the same group tend to behave similarly because they
have similar individual characteristics or face similar institutional environments – eg
individuals in the same school tend to achieve dimilarly because they are taught by the
same teachers.
Distinguishing between these are important. For example, if there are endogenous interactions,
increasing the academic performace will have positive effects on other students through feedback
effects. However, no feedback effects will occur if there are only contextual interactions or
correlated effects.
Matthew Robinson 41
8. Immigration
.8.1. Who decides to emigrate and where do they go?
From a destination country’s perspective, it is superior to have immigrants with high productivity
who adapt rapidly to conditions in the host country, thereby making a significant contribution to
economic growth.
Matthew Robinson 42
possibility of cohort effects. For example, if a study looks at waves of migrants entering the host
country in 1960, 1980 and 2000, the cohort may have different attributes which make it difficult to
estimate the benefits from assimilation; namely, if the cohorts have progressively less favourable
attributes, the most recent (2000) immigrants will have much lower earnings, compared to the
oldest (1960) immigrants – the earning differences could be purely attributed to assimilation, rather
than cohort effects.
• slower acquisition of host country skills: an ethnic enclave provides immigrants with less
exposure to natives and therefore hinder their ability to obtain necessary skills and
therefore jobs.
• Network effects: living in enclaves can increase the opportunites for employment because
the enclave disseminates valuable information (about jobs, etc), provides less
discrimination than the rest of the labour market. This therefore improves outcomes.
• Human capital externalities: The unskilled migrants may benefit from being in a group that
includes highly skilled members.
It is difficult to measure the effect of enclaves because selection is unlikely to be random, creating
a bias in results – ie if only unskilled workers go into enclaves, then outcomes may be deemed
negative.
Edin et al (2003) examins the effect of enclaves on assimilation in Sweden. They exploited a
Swedish government policy to randomly place refugees where housing was available (refugees
could leave, but 54% stayed in the same location for at least 8 years). The study found that the
being in an enclave had a positive, but not statisitically significant benefit. They found that those
with less than 10 years of education benefited from being in an enclave. They found that there was
a statistically insignificant result that highly educated people suffered from being in an enclave.
Matthew Robinson 43
Wage rigidity
In the case of wage rigidity, immigration increases the labour supply, which reduces the wage rate,
but not sufficiently to let the market clear. This means that unemployment rises amongst natives.
This results in employers gaining slightly. It is unclear whether the overall impact is positive or
negative.
• native workers changing their location due to the immigration – native workers may leave
the highly competitive labour markets, and/or there may be less domestic migration. For
example, the Miami population grew at 1.4% compared to 3.4% for the rest of Florida.
• Miami was extremely well set up to absorb Cuban immigrants – there were pre-existing
Cuban employment and social networks, local industries had demand for their skills and
the city had been accepting immigrants for decades.
Matthew Robinson 44
9. Labour market policies
.9.1. Types of labour market policies
Active policies aim to increase employment and wages of unemployed people, and include job
search assistance, training, employment subsidies and public sector job creation.
Passive policies aim to increase the material welfare of disadvantaged persons, and include
unemployment and disability benefits.
• Employers require a minimum skill level K, and the wage associated with this skill level is
W(K).
• A job seeker has a skill level of K*; the maximum salary the person can achieve is W(K*).
• Information is not perfect, so individuals do not know which firms require the skill set
encompassed by K*.
• Therefore, job seekers will apply to random firms, and will get offered the job where their
hiring standard is ≤ K*.
• However, the person will have a reservation WR, and they will not accept the job.
Under this model, the distribution (shown as a normal distribution) was centred around the the
weighted average of the offers between WR and W(K*). The tails are shaded out because they pay
too little or the required skill is too high.
The job-seeker’s reservation wage will respond to changes in the benefits and costs of search
activities as well as their discount rate. The reservation rate will be in at the intersection of the
marginal cost (upward sloping) and marginal revenue (downward sloping) curves. (dollars are on
the y-axis and wage is on the x-axis).
Workers with high discount rates have less to gain from additional searches, so the marginal
revenue curve shifts leftward, to a lower reservation wage.
Unemployment benefits reduce the marginal cost of the job-search, meaning that the marginal cost
curve shifts-rightward, increasing the reservation wage. Because the reservation wage has
increased, there are less jobs that are appropriate, meaning that the period of unemployment will
be longer.
Matthew Robinson 45
This arises because the costs of adjustment are sufficiently high to prevent movements to the new
industry.
(b.ii) Productivity
ALMPs may increase labour force productivity through either training programs or on-
the-job learning in the case of subsidised employment.
The productivity increase shifts the demand for jobs rightward, lifting employment
and wages.
Matthew Robinson 46
()c Deadweight losses and displacement and substitution effects
Subsidised employment programs are sometimes criticised because of:
• deadweight loss – a proportion of the people would have been employed anyway
• substitution effect – a proportion of the people will take the place of other people who
would have been employed in the absence of the subsidy.
• displacement effect – other employers may shed labour if they don’t qualify for the subsidy
to retain competitivemenss.
An OECD review of Australia’s JobStart scheme estimated the deadweight loss as between 67
and 79%.
Thus, wage subsidy and job-creation programs are generally considered to be high cost, low
benefit. But it comes down to a political decision as to whether the redistribution of opportunities to
a particular group are worth it.
• referral to work for the dole cause people to exit welfare payments when they would be
required to begin – governments hope this effect is minor.
The empirical study matched a treatment and control group by a number of factors that resulted in
people being able to be classified in one of 3,369,000 cells. It was possible to match 802 work for
the dole participants to control group individuals. The weighted average of cell differenes is taken
as the overall average effect of work for the dole participation.
Figures showed that more than a third of participants were in employment or education 3 months
after completion.
It did seem to have a positive effect on receipt of welfare payments.
Matthew Robinson 47
.9.5. Other empirical findings
Job broking:
• most effective where do not distrort the type of job search and where labour demand
conditions are most favourable
Training:
• effectiveness depends on ability to deliver skills that are in demand – if there is a poor
match, the lock-in effect is likely to dominate.
• least skilled individuals derive the most from these programs, probably because it gives
them the ability to acquire work skills.
Matthew Robinson 48
Economic growth (Jay)
10. Introduction
{not examinable}
At steady state, there is no movement from this equilibrium, unless one of these factors change.
Matthew Robinson 49
with depreciation.
They find that:
• 59% of cross-country variances in GDP per worker can be explained by the steady state
variables
• The magnitude of coefficients are too large – ie α = 0.6, implying that 60% of income goes
to capital.
Matthew Robinson 51
12. Easterly & Levine’s 5 stylised facts
.12.1. TFP is responsible for growth
Total factor productivity can be interpreted in a number of different ways, including externalities in
factor accumulation, adoption of lower costs of production and technological spill-overs. From a
regression point of view, TFP is basically the residual.
Matthew Robinson 52
.12.4. Economic activity is highly concentrated
Economic activity is concentrated. EL suggests that if there was no differences in TFP, factors
would spread out evenly. However, the fact that this does not happen may be due to technology
and externalities.
Matthew Robinson 53
Exam preparation
13. Sample questions from blackboard
.13.1. PART ONE
A PhD student is interested to explore whether state efficiency has an impact on tax compliance.
a) Discuss in detail what research methods the PhD student could use to investigate this
question.
b) What are the advantage and shortcomings of the suggested methods to answer this
research question?
c) The PhD student wonders whether it is possible to develop a theoretical model that allows
to explore the impact of state efficiency on tax compliance. Please develop such a model.
Matthew Robinson 54
My Essay
Introduction
The positive impacts that education has on the individual and economy are unquestioned.
Consequently, economic, social and political considerations necessitate a close analysis of
whether higher education is equally accessible to all Australian citizens. Australian evidence
suggests that low socio-economic status (“SES”) students are half as likely to participate in higher
education than their medium and high SES counterparts (“SES Inequality”) (James, 2002). Given
this inequality, eliminating SES Inequality to ensure an egalitarian and productive society,
constitutes a unique policy challenge for the Australian Government.
This essay establishes the existence of SES Inequality in Australia’s tertiary sector and why the
problem must be combated. It then briefly sets out policies that finance students’ university
attendance. The paper proceeds to a critical analysis of full tuition-fee scholarships, having regard
to its likely success in reducing SES Inequality. Finally, the authors consider alternative policies.
Australia’s education system and low SES disadvantage
More than one million people are currently undertaking study at one of Australia’s higher education
institutions, with more than 650,000 of these receiving government assistance towards their
education (Department of Education, Employment and Workplace Relations, 2008). Despite 25%
of the Australian population being low SES, over the preceding two decades low SES students
have consistently only made up 15% of the student cohort (James, 2002; Department of
Education, Employment and Workplace Relations, 2009). This indicates significant educational
disadvantage for low SES individuals.
SES Inequality is a legitimate ground for government intervention due to the economic and social
externalities generated by education. Higher education is a key factor in human capital
accumulation and is strongly tied to a nation’s economic growth (Jones, 2008). There is also
strong evidence linking education to increasing civil citizenship, reduced crime and increased
social mobility (McTaggart, Findlay, & Parkin, 2007).
Furthermore, there are a number of other private and public returns to higher education including,
inter alia, superior health, welfare, self esteem, parenting skills and social tolerance (Murray,
2009). Importantly, recent studies have found that the positive externalities associated with
education are greatest in the case of disadvantaged groups (including low SES), through
increased social cohesion and resilience (Murray, 2009). Therefore, it is clear that equal access to
higher education must be a high priority for policymakers.
SES Inequality has persisted in Australia despite the introduction of the first income contingent
loan system, the Higher Education Contribution Scheme (“HECS”), now the Higher Education Loan
Program (“HELP”) (Le & Miller, 2005). HECS-HELP is a sophisticated universal student financing
system, whereby the Government subsidies the tuition costs of students and imposes a student
contribution in the form of a contingent liability extracted through the tax system (Chapman &
Ryan, 2005). The student repays the debt once his or her income is above the prescribed
threshold. This means the risk is shared between the taxpayer and student. The success of
HECS-HELP spurred the United Kingdom and New Zealand, among others, to adopt income
contingent loan schemes (Chapman & Ryan, 2005).
Moreover, the existence of Youth Allowance and Austudy (“Income Support”) is important,
because it provides means-tested fortnightly payments to university students. Both Income
Support and HECS-HELP are seen to be essential to internalising the externalities generated by
higher education, and reducing credit constraints.
Australia’s policy settings provide a base level of income while at university that reduces the
quantum of forgone earnings, while deferring the tuition fees until individuals are benefiting from
tertiary education. These policies are incorporated into the model of university attendance below.
Matthew Robinson 55
Why do people educate themselves?
An individual’s income is predicted to increase by 7% for each year of higher education (Jones,
2008; Perna, 2005). However, this does not dictate that a person will pursue tertiary education.
People are assumed to treat education as a net present value (“NPV”) investment decision,
whereby a person will educate themselves where the marginal benefits exceeds marginal costs,
adjusted for their discount rate. The stylised model in Figure 1 can be adapted to take into account
the HECS-HELP and Income Support available to low SES students in Australia (see Figure 2).
This framework informs the remainder of the essay.
Figure 1: The costs and benefits of tertiary education over a lifetime
Figure 2: The costs and benefits of education for Australian low SES students
Matthew Robinson 56
The scholarship program considered
The essay assumes that the tuition scholarship program (“Scholarship Program”):
• an up-bringing that does not emphasise the benefits of education, leading to a short-
term focus; and
Matthew Robinson 57
2008). Thus, providing support to offset the initial costs of education is unlikely to be a cost
effective solution to SES Inequality due to its prohibitive cost and the small or non-existent
influence of credit constraints.
Cardak and Ryan (2009) find that ineligibility for university is correlated with a student’s SES.
Consequently, addressing the poor prior academic achievement of low SES students is a vital
policy consideration. The effects of low SES on education achievement are seen by Grade 9
(Cardak & Ryan, 2009), and may be evident as early as pre-school and primary school (Chowdry
et al 2010). Early interventions that:
• are ongoing;
• involve parents;
are posited to substantially improve non-cognitive skills that are essential to enabling students to
become eligible for university, counteracting the disadvantage suffered by low SES students
(Knudsen et al, 2006; Heckman, 2000). This is consistent with neurobiological evidence (Knudsen
et al, 2006) and the notion that learning begets learning (Heckman, 2000).
Further, structural changes to early, primary and secondary education is suggested as a cost
effective method of decreasing SES Inequality. Changes may include, increased transparency in
the public education system, introducing performance pay for teachers and increasing competition
between schools (Heckman, 2000). However, programs to reduce class sizes are unlikely to
return sufficient benefits to justify the substantial costs involved (Heckman, 2000). Other possible
policy responses include outreach programs between schools and universities, as well as
communicating of the benefits of education to low SES students and their families (Bradley et al,
2008; Milbourne & Bajada, 2008).
Finally, it is naïve to rule out non-institutional factors as a contributor to the problem. The family
unit, and their attitudes towards education, can have an important impact on students’ educational
development, aspirations and preferences (Heckman, 2000). Therefore, fostering an aspiration to
attend university and reducing perceived barriers to entry into university are further policy objective
that should be pursued to reduce SES Inequality (Dearden, McGranahan, & Sianesi, 2004).
Conclusion
It is clear that the alleviation of SES Inequality is a policy challenge, the resolution of which would
return significant positive externalities for Australia. However, the challenge cannot be solved by
implementing policies at the point of entry to university through tuition scholarships; if a small credit
constraint exists in Australia, it is unlikely to be solved by a Scholarship Program. Instead, a
holistic approach to education must be adopted, with a focus on the formative years of low SES
students’ education, promoting an aspiration to attend university and reducing perceived
inequalities.
Matthew Robinson 58