Tax revenue composition

- Social contribution: pension system (people who

work pay for pensions). Long-run problems due to
sustainability
23% - Corporate taxes: it’s a small share. They are paid by
33% Social contribution
multinational rms that although having large pro ts
Income taxes
don’t contribute that much. There is a tax competition
Taxes on consumption goods
Corp. taxes among countries ( scal paradises) to reduce corporate
19% tax to attract multinationals. Es. Ireland tried to get big
7% Others
multinationals by o ering nearly 1%/2% corp. tax.
19% - Tax on consumption (goods): VAT (value-added tax),
it’s the most important one
-Income and VAT tax have the same share in Italy. In
the US, most taxes are on income and the ones on
consumption are very small.
Public expenditure
Public expenditure expanded a lot after WWII. Now it’s about 50% of GDP but lately there was a decline:
- After 2000, many countries that relied on planned economy switched to a market economy (es. India or China)
- Body centre of the world is shifting from Atlantic (Europe) to Paci c (USA, China)
- Communist countries has switched to market economy

Theory (attitude of the market)

19th century was the golden age of market ideology. Great Britain was the champion of free trade. This ideology
founded the uni cation of Italy and Germany.
• 1850: Golden age
• 1873: Great Depression where people started thinking that free trade wasn’t the only solution seeing bank failures…
It a ected many countries simultaneously (France, USA)
• 1929: Other great depression where. Keynes developed microeconomics theories to solve the crisis.
• 1930-1970: great debate about which economic system is the best:
-> CAPITALIST (decentralized, only coordinated by market system, corporations make their own decisions)
-> COMMUNIST (planned economy: Soviet Union did not catch up with the new technologies and development).
It was a debate without an end. Historical experience provided a solution to planned economy since after 1968
innovation made the di erence as planned economy wasn’t longer able to keep up with innovation pace,
sophisticated technology and capitalistic economy.
• 1980s: beginning of the process of capitalisation and liberation. This had a huge impact and 500 million people
were saved from poverty. Es. China moved to capitalism.
Then there was a demise of the industrial policy: the government
Nowadays there is the idea that if you want to compete on the market you need to imply protectionists measures
implemented by the government.

Japan 1950-1960 approach then demise of industrial policy: industrial policies and corporations guided from above,
Japan investments made by private entities but the government will guide them.

Over the last 10 years, even if production and innovation is provided by privates, the idea that governments must
intervene spread leading to the need of need for stronger intervention. (EU, USA).
This is a way of imitating china but in a minor extent.
In the 50’s there was a huge debate about capitalism yc planned economy (some thinks this is not working)
ff
fi
fi
fi
ff
ff
fi
fi
 1. MARKETS AND EFFICIENCY
1ST FUNDAMENTAL THEOREM OF WELFARE ECONOMICS: ARROW-DEBREU’S THEORY
Competitive markets (decentralised) are e cient.
Microeconomics is based on 3 concepts:
1. Rationality of agents
2. Equilibrium markets equilibrium: demand=supply
3. E ciency: a way of assessing the market through “good” or “bad” equilibrium. His idea was that “an outcome is
e cient if there is no other feasible outcome in which everybody is better o . The founding assumption is that the
size of the community is xed.
Assumptions:
1. Individualistic (assess individuals one by one) ≠ collectivity
2. Welfaristic: compares outcomes based on welfare ≠ liberty (exercised by saying yes/no) removing the ethical part
3. Consumer sovereignty: consumers know their preferences.

Pareto e ciency theory (An outcome is e cient (A) if there is no other e cient outcome (B)) has a very weak
requirement because it is not selective enough.
Between two outcomes which are Pareto e cient I should use which criteria?

Ex: single good Pareto is mute

In titanic tragedy 2000 people died out of 3000. Most of the people who died were relatively poor.
After that accident a regulation was made to cover everybody. Who should be saved?
Major ethical issue: only one valuable good life
1. Youngest ( years of life saved)
2. Random (risk equity)
3. Socially valuable

It is very di cult to decide. Class voting chose 2.

Pareto would say “choose as you like as long as the life boats are occupied”
Pareto + ethical criteria is the solution, in order for Pareto to work you need two valuable goods
In titanic case there are two valuable goods: lives and money
Who pays to be saved and who receives money?

EDGEWORTH BOX
Simple box in which 2 economic system are analysed.
Simplest possible market: 2 individuals and 2 goods
The solution is to give A dogs and B cats

Preference is converted into terms trade:

- Individual A is a cat lover
- Individual B is a dog lover: if you take away 1 dog, it has to be compensated with 2 cats
Individual B wants to nd best point given the terms of trade, how many cats can he have by giving away dogs?
Individual A wants to nd best point given the terms of trade, how many dogs can he have by giving away cats?
The marginal rate of substitution are di erent.

In order to nd e cient allocation you combine the budget lines of the individuals.
The endowment of cats and dogs is xed.
Side of the boxes de ned by total amount C̄ D

A gives away all his dog moving to a new indi erence curve, to be better
B gives away all cats.

A and B are allowed to trade and they move to another outcome where they are better o .
ffi
ffi
ffi
ffi
fi
ffi
fi
fi
fi
fi
fi
ff
ffi
ffi
ffi
ff
ffi
ff
ff
ff
 General equilibrium theory: under weak condition trades happen
Basic condition/ assumptions:
1. Individual rationality: market is based on free will and voluntary, people aren’t forced and cannot be worse o .
2. There is an auctioneer, who is an external individual that sets the price. People are price-takers.
es. In the nancial markets the terms of trade are set by an algorithm that sets the price. People observe the price
and decide what to do (buy/sell)

E cient allocations
Having only 1 good, Pareto e ciency has no meaning and isn’t selective.
With 2 good Pareto helps: an allocation is e cient if there is no other allocation where everyone is better o .
“Feasible” means that the allocation is compatible with resources available in the economy (in the box)

E1 isn’t Pareto e cient

On E2 both individuals are better o , being on a higher indi erence curve.
On E3, A only has cats and B has all dogs (+some cats). This situation is
e cient because you can’t improve from it.

The set of e cient allocation is on the border

E ciency is about ruling out the wasteful allocations, not compatible with
individual preferences.

To choose among the e cient allocations, Pareto criteria must be supplemented by an equity criterion.
Which is the most equitable allocation? The Pareto criteria is only the very rst step.

M= market outcome
E= starting point people who are born in poor/rich families
Through trade you end up to a nal outcome, through the market that
increases the wellbeing of people.

There is a constraint given by the fact that people make free choices and
cannot be worse o .
ffi
ffi
ffi
fi
ffi
ffi
ff
ffi
ffi
fi
ff
ffi
ff
fi
ff
ff
 2ND FUNDAMENTAL THEOREM OF WELFARE ECONOMICS: ARROW-DEBREU’S THEORY
All e cient allocation can be market outcomes if you can redistribute resources.
E cient allocations can be obtained as market outcomes redistributing resources and change starting point.

From E to E1= redistribution, it can be done by somebody who has the power/force to
take away the resources from individuals. Market works bringing to another outcome.
The market takes to one allocation that might be unfair (someone is richer or poorer).
Market mechanism is supplemented by a conditional mechanism= taxes
(redistribution).

The state can choose the most equitable allocation

Market+taxes (redistribution) can get to any e cient allocation (Without considering the fair ones)
Bottom line of this geometrical analysis= equity and e ciency are compatible in a market economy, with taxes.

Back to market analysis

Redistribution is a political issue:
- egalitarian societies (Es. Scandinavian countries)
- Less egalitarian societies (Es. In USA there are a lot of di erences in terms of economics levels of well being)
Income taxes touch the issue of equity: they are based on the idea that the rich should be taxed more of the poor.

SURPLUS ANALYSIS
Assumptions:
- the goods are money and dogs.
- Assumption: money has no satiation (the more money the better)
- Preference for dogs is independent of how much money you have, the intensity of preference is the same
Money as a measure of surplus: from general equilibrium theory to a setup in which one of the goods in the
economy measures the surplus people get from trade

Example 1: Titanic
2 goods: lives and money (people care about both)
People can be paid to give up their place on the boat or can be taxed to get the place.
Pareto e cient allocation: it is a measure of sel shness or others-oriented attitude of people.
Who should be saved? People who have the highest willingness to pay, express the highest “demand”.
How much money is an individual willing to trade to have a place on the boat.
Market outcome (Pareto e cient):
• People with the highest willingness to pay get compensated
• Receive money to drown
Implementation—> auction
This outcome is not e cient, as e ciency must be associated with a tax system (according to equity criterion). Before
the auction starts, income must be redistributed.

Example 2
Consider a valuable item in the economy: a piece of land and there are potential buyers (2 individuals: school and a
rm). Who should get the land?
The school values the land 100, while the rm’s value is 120. The e cient outcome is for the good to be allocated to
the rm since the willingness of the rm to pay is higher. According to equity (e ciency+tax): the land should be
assigned to the rm for 120, the rm makes a surplus of 120, the rm can be taxed and the money can be transferred
to the school. There is a tax that redistributes the gains from trade.

Consumer and producer surplus in a market

Statistical econometrics technique to estimate demand and supply.

Supply captures the marginal cost of producing the product.

Demand captures the willingness to pay (marginal bene t)

Total welfare= CS+PS

Competitive equilibrium maximises total welfare —> Q*=W(PS+CS)
fi
ffi
fi
ffi
ffi
fi
ffi
ffi
fi
ffi
fi
fi
ffi
fi
ffi
ff
fi
fi
ffi
ffi
 E ciency has 3 dimensions
1. Allocative e ciency: goods one produced are distributed in an e cient way, i.e. the e cient way is to
assignthem to the ones who value them the most (allocate according to willingness to pay)

Pm= price set by a monopoly, which allows to get decent income.

Even if it is price- maximising, it is ine cient: there are people with high willingness
to pay that cannot a ord the good and there are goods not sold.
A part of the population does not take the service. However, it is not pro t-
maximiser for the producer.

2. Production e ciency: all the items are produced by the rm that produces at the least cost

Supply curve is built at the level where MC is at its lowest price and larger
amount. Firms that can’t sustain these prices are forced to exit the market.

3. The match between D-S: what consumers want and rms can produce giving the optimal quantity. How much
units should be produced such that MB>MC. Marginal costs should be lower than the bene ts.
X: the value the consumer attach to the item (b) is less than how much it costs tot he rm to produce it

Monopoly
In a monopoly, prices maximise pro ts but there is an ine ciency= DWL,
set of people who value the good at a price which exceeds industrial
production, but those units are not produced

Ine ciency due to high prices

—> Consumer gained and producer loose

Example 1
A representative for taxi drivers states the price is reasonable (price based on an algorithm between prices and km)
-> How can we argue this is not good? We argue about price: the price is set by local government and not by the
market, thus it’s ine cient because prices are too high and people with a high willingness to pay are not served.
-> Which could be a counter argument? A political argument
Example 2: rent market
ffi
ffi
ffi
ffi
ffi
ff
fi
ffi
fi
ffi
fi
ffi
fi
fi
ffi
fi
 Ine ciency due to low prices
Example: Rent control
In the 70s the rent price was set by the government to avoid exceeding some amount. This, however, has some
issues. Given the price of P upper bar, the supply would only bring to the market Q upper-bar which is the new
quantity in the market. Suppliers cannot supply more than they want.
Consumer surplus would become larger and producer surplus smaller.
In general:
- if prices are higher, more apartments should be available on the market
- If prices are lower, less apartments should be available

Economist would argue the outcome is ine cient: there a lot of apartments people will be willing to rent at a very
high price. Also, the landlord would be willing to rent a very low opportunity cost. There would be man opportunities
to grade and make both sides better o .

There is a tax system: redistribution through which renters receive an endowment, but still paying a competitive price
Direct subsidies help the renters without interfering the market (reducing the overall amount of surplus)

E ciency outcomes given 2 di erent scenarios:

1) The government can’t use tax system
- Cross countries, international relations, each country cares about its share of the surplus. An agreement/measure
is always ine cient. The total surplus is not maximal, every country cares about its own share. Without direct
compensation, it makes no sense to talk about e ciency because there is nob redistribution
- Developing countries with weak tax administration. Redistribution wouldn’t take place after maximising surplus

Pm= monopoly price, which is the price that maximises producer surplus
Pc= competitive price
Ps= monopsony price, which is the price that maximizes consumer surplus

What happens to consumer and producer surplus if price goes from zero to highest level:

- Getting to Ps, both consumer and procure surplus increase, but consumer
surplus is as large as possible.
- In Pc all surplus are high
- Getting to Pm, the producer surplus is as large as possible

Without taxes, thus redistributing resources only through P, prices between Pm and Ps are e cient.
Prices below Ps are ine cient because there is another price, which is higher, where both sides are better o .
reaching Ps, CS is as large as possible. The same applies to the prices until Pm, where Pm is as large as possible.
Without a tax system, according to Pareto extreme prices should be avoided.

2) Government can use tax system

Market creates as much a possible surplus and taxes redistribute. W=CS+PS

Resources can be redistributed from one side to the other

Pm is ine cient because there is another situation in which both sides are better
o . Moving to Pc (competitive outcome), resources can be redistributed taking
away 1 euro from the consumer to the producer. In B both sides are better o .
ff
ffi
ffi
ffi
ffi
ffi
ff
ff
ffi
ffi
ff
ffi
ff
 Variation: costly tax system
Transferring resources from the competitive outcome, a share of them gets wasted.

e.g. a share of the euro that would be given from the consumer to the producer gets lost.
Only 0.8% of the euro goes to the producer (“leaky bucket”). Transfer happens through a
steep line, not 45 degrees line.
In his case, the set of the Pareto e cient prices is larger.
The more leaky the bucket, the more costly the transfer, the larger the e cient outcomes

Cost of redistribution
Cost = 80% (super conservative), it’s too costly to redistribute
Cost= 5% super progressive
Cost of the tax system includes:
1. Cost of administration (100.000 hired people in Italy )
2. Compliance costs (formalities, es. Bolla di accompagnamento)
3. Disincentives (impact on people after taxes)

MARKET FAILURES
Conditions under which the market is ine cient (market economy isn’t able to deliver an e cient outcome).
1. Market power (monopoly)= causes ine ciency. It leaves consumer not served. (implications studied in
competition economics, industrial organization)

2. Asymmetric information: some parties have better information (di erent from equally ignorant). The market is
e cient in dealing with uncertainty.
Field developed from 1970s onwards (Ackerlof) making huge di erence.
Example: adverse selection model. Considering market for “lemons” (American slang meaning for defective
product) and market for “Prune” (market for goods products). Going to a park dealer in the US trying to buy a
second car, there is a possible distribution of quality of the case (from 0 to 1). In extreme situation the buyer
doesn’t know anything about the quality of the car while the dealer knows everything exactly. In an uncertain
situation the price would be 1/2. Supposing the seller knows the quality is worse than 1/2, if the buyer asks for this
value, the dealer will give him a car whose value >/= 1/2. The actual value would be 1/4… any price proposed by
the seller would give him a car with a lower value. The only price that makes sense is 0 (If I propose something I
will only get something which is worse). The market completely collapses and only lemons survive.
If there are asymmetric information, usually there are a lot of policies (mandatory certi cations, mandatory audits,
minimum quality standards, regulations) which aim to reduce it

3. Public goods: are not rival in consumption (many people can bene t from good at the same time)

4. Externalities: side e ects of economic activity (ex. pollution, carbon emissions) +

5. Missing markets: some tools are not traded in market equilibrium

Ex.There are 4 goods (x1,x2,x3,x4) and there is a price for each apart from the last one. This market is not
e cient and this a ects all the other prices. In order to be e cient, the market must cover all the possible
commodities.
ffi
ffi
ff
ff
ffi
ffi
ffi
ffi
ff
ff
fi
ffi
fi
ffi
 2. PUBLIC GOODS
Futures of public goods:
1) Non-rivalry: more people can bene t at the same time.
Examples: lighthouse, bridges 1(9th-century), public defence, very huge topic.
In these cases, the cost of adding one consumer is zero.
Congestible goods= marginal cost of adding 1 person becomes larger as number of people increase (es. bridges
can be congested, turning from non-rival to rival).
2) Non-exclusivity: all people can bene t from the service and they cannot be excluded from consumption.

NON RIVALRY
This is a competitive market failure.
Example1: suppose setting up a market for a lecture, the cost is 20$. A competitive market applies marginal cost.
Marginal cost of serving 1 additional person is 0. Pricing lectures would exceed marginal cost, excluding consumers.
This would create distortion which is a fail (similar to monopoly) as public goods should cost nothing in a competitive
market. This explains why non-rival good cannot be provided by the market as the e cient price would be 0.
Non exclusivity is an additional problem.
Classic solution= public provision, provision isn’t about the production process but who pays for the good.
Usually the state uses public money (tax revenue to provide to this good). Tax payer nances the good.
Example 2: Production of knowledge.
Einstein formula is a major discovery that expands knowledge and anybody can bene t from it (non-rival good).
Most scientists are public employees which have a lot of public support. (example of classic solution )
Research in Europe is 50% public and 50% private.
Example 3: Production of music (non-rival good).
A musician produces music and anybody can bene t from it, once produced.
Classic solution: musician is made a public employee paid by the state
There could be a problem associated with this, problems:
1. Real value: it’s hard to estimate the value of service.
2. Real costs: what is the value of music
3. Potential mismatch between those who pay and those who bene t.

Switching to a market solution you can solve this problem

Traditional solution might entail ine ciencies too and this is why we have both public and private solution.
Resolution: 50% public (basic research) 50% private (applied research, having industrial application)
Pharmaceutical research is nearly 90% private.
Pharmaceutical product are going to produce ine ciency because pharmaceutical drugs once you have the formula
is a non-rival good. These products are provided by private market, meaning they will be provided ine ciently
(people will be excluded). Drugs are produced by private markets thanks to protection IPR, which entails huge
welfare costs.
In the private sector good become exclusive and this choice is a political/ policy choice.

SAMUELSON RULE (EFFICIENT LEVEL)

Under the Samuelson rule, the e cient quantity of public good Q should maximize:
n

∑
W (Q) = = b′(Q) − c′(Q) = 0
i=1
n
That is Q* should solve ∑ = b′i (Q *) = c′(Q *) (some of the marginal bene t should equal the marginal cost)
i=0
n
• This formula maximizes social welfare (W (Q ) = ∑ ) when the size of the community (n) is xed.
i=0
• The bene t from the public good is measured on each individual which bene ts from the consumption.
n
• W= ∑ =b’(Q)= social marginal bene t for people in the community, which are decreasing.
i=1
• c’(Q)= cost of production is non-decreasing, independent of the good and how many people are served

Formula says that you have to produce quantity up to the point in which the sum of the marginal bene t equals
marginal costs. (How many units of public goods should e ciently be produced? this works if there is redistribution)
Not related to nancing (no concerns about who pays), but to the e cient amount to be produced.
n
Variation of the rule: ∑ M R S i = M R T
i=0
Marginal rate if substitution (public) equals the marginal rate of transformation (private).
It applies when there are no rules in the economy.




fi
fi
ffi
ffi
fi
fi
fi
ffi
fi
ffi
fi
ffi
fi
fi
ffi
fi
fi
fi
ffi
fi
 Example: How many reworks should be shouted as they are very expensive.
1. Everybody is allowed to see them (non rival good) cost
2. Samuelson rule (sum of marg. bene ts= marg. costs), it’s about balancing bene ts and costs.

What happens to Q if n increases (if more people want to see the reworks, should there be more/less reworks).
As this is a non-rival good, the cost is independent from number of people.
If new people like reworks, e cient Q increases.
n
∑
=b’(Q) , if Q increases, the sum of the marginal bene t increases.
i=1
Dog owners have negative marginal bene t from reworks because dogs get scared. If most of the new people are
dog owners (with negative evaluations for them) the rework are harmful and Q decreases.

It’s hard to how people value the public good ( nd out bi’ and c’). This is done through
1. Surveys, to get an estimate
2. Indirect methods, to measure how much people like public goods (transpiration cost)
Ex. 1: usually applied to natural parks. Which is the distance they are willing to make. The average distance tells
is an indirect way to calculate the bene t people get from getting to the park.
Ex. 2: Edonic pricing: there is a park and you check in the neighborhood all the house that are close and price of
houses. If the price increases, it means a good service is provided by park making the area more valuable.
3. Referenda, very coarse and bad info since it’s s binary (yes or no we can’t get the detailed information)

————————————————————————————————————————————————————-
Under perfect information, the government could collect taxes equal to the bene t of individuals.
So, each individual would pay at tax ti (Q*) = bi (Q*)

If the government does not know individual bene t functions, we are in trouble. If you tell people that they will pay a
tax equal to the bene t they net for the public good, they will tend to under-report their bene t (so as to pay less).

1. Vicktrey Mechanism
Vickrey found a clever outcome to overcome the problem of preference misreporting: idea that people’s payment
should not be directly tied to the bene t they report.
Starting with Q0 of public good, you want to know whether Q should be increased.
n
Knowing the marginal bene t you would decide to increase the price of public good if ∑ b′i (Q0 ) > c′(Q0 )
i=1
Yet b′i (Q0 ) is unknown-> mechanism according to which you take individual k and ask him what the marginal bene t is.

Individual K becomes pivotal and doesn’t gain by misrepresenting his preferences.

Suppose that b′k (Q0 ) > c′(Q0 ) − ∑ b′i (Q0 ):

i≠k
If the marginal bene t reported by k is greater than c′(Q0 ) − ∑ b′i (Q0 ) , she gains b′k (Q0 ) and she pays c′(Q0 ) − ∑ b′i (Q0 ).
i≠k i≠k
The gain exceeds the payment only if
• b′k (Q0 ) > c′(Q0 ) − ∑ b′i (Q0 )
i≠k
n
∑
• b′(Q0 ) > c′(Q0 )
i=1

Thus the additional unit is produced only if it’ e cient to do so.

The mechanism applies this scheme to all players and you end up with the e cient level of the public good (although
not necessarily with enough money to nance it ).
















fi
fi
fi
fi
fi
ffi
fi
fi
fi
fi
fi
ffi
fi
fi
fi
fi
fi
fi
ffi
fi
fi
fi
fi
fi
 2. Voluntarily contribution
The public authority doesn’t decide. In this case every individual decides simultaneously how much to contribute to a
common pool (common fund) of money T for their own bene t and how much they value the service. Q is provided
depending on how much money is raised. This amount of money is transformed into a pubic good acceding to the
production function Q = f (T ) (with T euro, cost, you get f(Q) unties of public good). Each individual is endowed with
a xed amount of money Mi, es. monthly salary.
This amount of money can be used for private consumption mi or the contribution to the public good ti: Mi = mi + ti:

Individual welfare is
wi = mi + bi(Q)
wi = Mi − ti + bi(Q)
- Mi − ti = private consumption
- bi(Q)= what individual bene t from public good, which depends on wi= welfare level

Where Q = f (T ) = f (t1 + t2 . . . tn )

The optimal contribution for individual i solves

At the individual optimum: b′i (Q)f′(T ) = 1 (marginal bene t)

• one euro of contribution produces an additional amount of public good that yields a marginal bene t of one euro
• Agents considers the bene t from the public good

This contribution is not e cient (and low).

How much does overall welfare change if ti increases? How much should individual contribute to the pool?
For e ciency, contribution ti should maximise social/ total welfare (W= w1+w2+w3…wn)

--1= individual is going to su er giving 1 € up

n

- ∑ = b′k (Q)f ′(T )=> all other individual bene t

k=1

Thus individual private bene t = individual private cost

E cient quantity:
n

∑
= b′k (Q)f ′(T ) = 1
k=1

In the special case in which all individuals are the same, the social optimums is given by
n b′(Q)f ′(T ) = 1
where the social bene t is n times larger than the private bene t

• one euro of contribution (by individual i) should yield a social bene t equal to one euro.
• E cient amount is calculated by considering the bene t for all individuals in the society








ffi
ffi
fi
ffi
fi
ffi
fi
ff
fi
fi
fi
fi
fi
fi
fi
fi
fi
 Since the social bene t is larger than private bene t, e cient quantity is larger than quantity resulting from voluntary
contributions. This proves that by relying on voluntary contribution, it’s possible to collect money that is less than the
e cient level (T<T*= e ciency level). When people contribute they only take into account the private bene t.

In the special case in which all individuals are the same, the social optimums is given by
n b′(Q)f′(T ) = 1
where the social bene t is n times larger than the private bene t

Cost-bene t analysis
Climate change is a global issue: it’s a public bad as it negatively a ects everybody.
Main issue: is who pays for it.
-> Here there would be a voluntary contribution game.
People bene t from a reduction in climate change but at the same time, it’s costly. If es. Italy moves towards this
transition there is private bene t + huge global bene t. Should Italian politicians consider that whole world bene ts
from transition since cost is private and bene t global (free riding).
In the US social cost of carbon is an important measure: it’s estimated by economists considering what is the
adverse impact on climate change and the cost for the environment/ societies caused by 1 ton of CO2.
Under Obama this cost =50-300$ (cost if each ton of emission of carbon having for the world impact)
Environmental agencies stated the social costs of carbon 48$.
Trump used a di erent estimate for the social cost of carbon, becoming 5$.
To compute thus calculation, , the bene t of US economy should be considered, not the global one. This has
implications changing all policies assessement.

3. Local public goods (Tiebout)

Given a community in which individual are more or less similar, public good nanced by the state require taxes.
People have di erent preferences: some of them are willing to pay high taxes as they want a lot of public goods and
other people want less public goods and less taxes.
Solutions:
- In America’s federal system (50 states) public goods and taxes provisions are provided on a local level.
“Voting with your feet”= people move to di erent states providing the best combination of taxes and public goods
Across jurisdictions there is competition to attract more people.
- In Europe there is competition between countries but mobility is limited.

According to Tiebout, people are encouraged to move to nd the best solution

E cient provision of public good

1) Authority/government could impose the solution (how much individual should contribute)

Talking about di erence countries (es. China or US),

2) Binding agreement: country A contributes x and B contributes Y. Countries can rely on their own obligation and the
n
contract. If people could make this agreement, toutcome would be e cient ( ∑ b′i (Q ) = c′(Q )), such that individual 1
i=1
pays t1, ind 2 pays t2 and so on.
If they cannot everyone decides to the base of their own interests getting to the ine cient outcome.
E cient outcome is possible but requires some conditions.




ffi
ffi
ffi
fi
fi
ff
ff
ff
fi
fi
ffi
fi
fi
ff
fi
fi
fi
ffi
fi
fi
ff
ffi
fi
ffi
fi
fi
 3. EXTERNALITY
Economic phenomena and side e ects not mediated/captured by market mechanism.
In order for externalities to make sense there must be an initial general equilibrium.
If A has an impact on B and is not mediated by general equilibrium= externalities.
It must be clear what belongs to the framework.
Q* is not e cient anymore if it’s a ected by externalities.

Positive externalities-> positive side e ects, externalities provide bene ts

Negative externalities-> harmful side e ects

How externalities a ect the equilibrium on the market?

Example: Consider negative externality produced by car manufactures corresponding to toxic emission having
adverse e ect on population. In this case externalities are measured in € (usually the hard part is the conversion).

Externality is proportional to the quantity produced-> each unit of car (Q)

produced creates e.
Total externality, total harm= E
Initial supply curve accounts for manufacturer costs.
Supply curve shifts upwards and includes marginal social cost.
Social cost includes manufacturer cost + externality.
Starting form Q*: bene t of the consumer to consume one additional unit.
The cost to society is the manufacturer cost and e (additional cost on society).

E cient production level

Q* is not e cient anymore because Q*: e+c’>b’ (social costs of additional unit exceeds the bene t)
Quantity should be reduced by m. The externality is not mediated by the market

Qe : e+c’=b’

Deadweight loss
DWL is associate with the units between units from Qe to Q* (externalities in excess)
Overall level of externality (E) exceeds the bene t (b’) => SC (c’+e) > B

• DWL increases as we move far away from equilibrium

• One of the edges of the triangle is at the e cient point (here DWL=0).
• DWL= social cost - surplus
• Surplus= bene t of market participants (consumer and producer)

SW = CS +PS -E
Social welfare= consumer surplus+ producer surplus - ext
It’s important to know the social welfare to then redistribute the money.

Size of the DWL

It is determined by the elasticity of demand and supply

DWL triangles have same height but one is short and the other long.
DWL when supply and demand are elastic is bigger.
ffi
ff
ffi
ffi
fi
ff
fi
ff
ff
ff
ff
ffi
fi
fi
fi
PIGOUVIAN SOLUTIONS
Methods to reduce/address the externality developed by A. Pigou.

(1) Regulation (administrative law in Italy). It is a huge field, not very developed in common law countries.
It means that a policy maker/ authority gives rules that can be very expensive and can touch any activity.
Example: what technology can be used? All harm coming with this product was approved since permission was
previously asked. Which products can be marketed? Electric vs traditional cars.

Among many regulations: government says

• at most Q cars can be produced (Q<Qe)
• what is the amount people are allowed to consume or produce
• what is the allocation

(2) Corrective taxes (command and control system). The market shouldn’t be interrupted but there should be a
tax, discouraging consumers and getting to an efficient outcome.
Producers internalize the externalities paying for the damage they cause (measured in €). If taxes is too high this
leads to inefficient outcome.
Example: Any time you produce car you pay a tax (t)= e.

Tax revenue= Q’ x t
tax improves Economy and brings money to the government

Correction applied to postive externality

SC (manufacture cost - e) of the activity is lower. Efficient quantity is greater than market point, thus production
isn’t enough. Government can impose a subsidy to increase the quantity of the market (anytime you do something
good you get money). Through the subsidy the market shifts to the efficient point. DWL should be proportional to
excessive quantity

Economists think tax is the best tool but from an efficiency point it makes no difference.
Pigouvian taxes are relatively rare.
Example: tax on tobacco or tax on oil (oil consume leads to a lot of toxic emissions. 50% of the price paid on oil is
tax payed to the government)

(3) Liability law: idea that if you cause harm you have to pay compensation for damages. This would solve the
externality as a tax but the revenue goes into the end of the victims and not the government.
RONALD COASE
1. Coase conjecture
2. Externalities and Coase
3. Example (Coase vs Pigou)
4. Problems of coasian solution

1. COASE CONJECTURE
Monopoly set an initial price c1 and a very high later p1 creating a distortion.
Coase argues that most goods are durable (they last es. washing machine).
From a dynamic perspective, monopolies have incentives to discriminate so
that people that have an urgency to get it buy it immediately and over time
the same item will be sold at lower price.
Thus with durable goods, there is price discrimination and no inefficiency.
His aim was to fight market failures.

2. EXTERNALITIES AND COASE

Ronald Coase wanted to fight Pigou.
According to Pigou, the state would’ve to intervene in case of externality (idea
argued by pro-market people)
Coase proposed a change in perspective:
The externality is not a product of the market but is due to a lack of market.
Victims aren’t included in the market and they should have the possibility to contract with the “polluter" (who
creates the externality) and they can either:
- pay the polluter not to pollute
the polluter has been authorised to carry out its activity and he knows its activity has some harmful effects
- Be paid by the polluter not to pollute.
This activity isn’t lawful but still is so profitable that owner pays people to authorise it.
In this situation the entitlements are different.
For Coase, the first decision is who can do what= entitlements (es. Right o pollute, right to clean air), than allow
people to negotiate. By carrying out negotiations you create a market and reach efficiency solution.
Coase argues that the solution is independent from entitlements.
Solution: more market transactions (not the state)

3. EXAMPLE (PIGOU VS COASE)

There are 2 possible ways to carry out an activity:
- high production level: high profit πH=100 and high externally damaging for population EH=80 (WH=20)
- Low production level : low profit πL=70 and low externally damaging for population EL=20 (WL=50)

Efficient production level is computed calculating social welfare level WH. From an efficient perspective low
production level WL is the efficient one. This can be done if there is a tax system.

-> Pigou solution

State should regulate the market (“tame the market”) as it doesn’t work properly
State should intervene through:
1) Regulations: “command and control”.
The simplest case is to impose low level of production. Producing too much is forbidden Qmax=70.
all market system is subject to political power.
2) Corrective tax: tax for the producer is equal to the externality.
If level of production is high tH=80 or if you produce little tH=20. Profit of the firm becomes lower:
πH=100-80=20 and πL=70-20=50. The firm decides level of production influenced by the tax.
3) Liability law: if you cause damages you have to pay.
lH=80 and lL=20 are thew payment for damages to victims. Profit become πH=100-80=20 and πL=70-20=50.
Firm internalise the externality and voluntary decides to undertake low production level.

At European level-> principle of environmental law

“Polluter pays principle” (applies only to environmental damages, to natural assets).
-> Coase solution
He doesn’t want state intervention, people have to be provided with entitlements (rights).

2 possibilities:
a) Victims are entitled to clean air (no pollution).
This is an absolute right, if firm pollutes (E>0) it has to close down. BUT parties can negotiate. Considering high
level of production, the firm is allowed to produce creating welfare. If victims get 100, they allow the firm to carry
out activity and pollute. Victims are paid 100 and net 100-80=20 or they are paid 70 and net 70-20=50. This is an
extreme where victims have all bargaining power.
b) Firm is entitled to pollute as much as it wants (E-> ∞).
Victims pay money not to pollute.
H: victims say if you don’t create externality EH=80 we pay you 80
L: victims say if you don’t create externality LH=20 we pay you 20
Maximum pollution is 80. If E<80, firm get money form victims.
Firm make money by not polluting:
H: πH=100+0 (doesn’t get any money because E is max)
L: πL=70+60=120 (firms receives 80-20=60)
Firm makes more money by polluting less and receiving more compensation from victims, internalising E.

Both a) and b) lead to an efficient solution where externality is internalised.

=> NO inefficiency if people can trade on the externality (people negotiate about the level of the externality)
Allocation of the entitlement doesn’t matter.
The important thing is who is the seller and who is the buyer and that people can negotiate.

If externality persists, there is a problem in the negotiation state and state should give more facilities in trade.
In reality negotiation is hampered by transaction cost (write/enforce contract, hire lawyer, find the other party…)
Role of the state: facilitate these negotiations by reducing transaction costs and empowering the market

Caveat: from a distribution perspective, allocation of entitlements matters a lot.

Victims are better off if they get the endowment (= right to clean air) rather than buying endowment by someone
(If I am the victim I prefer to be entitled clean air and get compensated, rather than paying)
This distributional issue has nothing to do with efficiency, but about equity (how income is distributed).

4. PROBLEMS OF COASIAN SOLUTION

Coase is very pro-market and his idea not always works (problems= transaction costs).

Example 1: pollution is so large that health issues outweigh the pro t. The rm has the right to pollute. If people
cannot coordinate to come together and negotiate to the polluter.
Example 2 “calabres": in the roads there are sidewalks, pedestrians and cars. Many car owners and pedestrians want
to use the road. Car drivers who want to use the road are potentially dangerous and would have to negotiate with all
pedestrians. To get to a coasian solution either pedestrian are entitled to a car free road where car owners pay pr
other way around, which is unfeasible.
Solution: instead of asking the party to negotiate with themselves, the state nd the price for the risk. If accident
occurs, party has to pay damages.
This brings to the 3rd Pigouvian solution
In the cousin solution, people have property rights and have to negotiate ex-ante if they want to undertake dangerous
activities while with the pigouvian solution the state decides that if you harm people you have to pay damages,
deciding the price.
fi
fi
fi
 EXTERNALITY AND EMISSION
Since 1970 climate change revolves around the question of CO2 emissions and equivalents, and greenhouse e ect.
At the beginning it was hard to understand that CO2, a natural element, can be dangerous.
Goal-> reduction of emissions
In EU there’s a commitment
- - 40% of emissions written in 1990 by 2030
- Carbon neutrality by 2050 (no new CO2 in the atmosphere)

Previously, in a competitive market, to reduce externality, quantity in the market Q* is reduced to e cient level QE
These diagrams focuses directly on reduction of emissions (not quantity).
The more you invest and reduce emission, the higher the cost but the higher the bene t.
To get to the e cient level of reduction R*-> maximise the di erence between marginal bene t and marginal cost

Adapt diagram to Pigouvian solution:

1st solution: Regulations
Regulation state to reduce emission by es. 40% by installing lters.
Problem: asymmetric information about the cost (c’)
There is not a uniform cost for all rm a ected even and di erent rms are a ected by same regulations
A has large marginal cost while B has smaller marginal cost.
Imposing same rule on both rms leads to ine ciency. It’s possible to measure ine ciency

• For A, marginal cost of the regulation is 100. A has to reduce

emission such the cost of one ton reduced is 100.
• B reduces emission up to the point where one ton of reduction
costs 50 euro.
Cost of A> B, so B could bargain saying it can reduce more if it
is paid 100 and A reduces one less saving 100.

Firms facing lower marginal cost can compensate the other by lowering emissions in a more e cient way.
They can get lower amount of emission in a cheaper way.
Market is e cient in 3 dimensions: between these production is carried out by the rm which is cheapest cost
producer. With a uniform regulation, each one produces the same amount creating ine ciency that can be measured
looking of the cost of compliance of each rm.

2nd solution: Tax

If rms have di erent marginal costs, a tax can be imposed to reach an e cient outcome at an aggregate level.
Tax measures W= marginal bene t for the rm of reduction.
Emission tax: for each ton emitted you pay 100 euros. Increasing reduction of 1 ton you gain 100.

Firm A goes to a di erent optimal solution R*.

By raising or lowering the tax level of emission changes,
caring only the aggregate level (tax up and R grows).
If marginal cost is low, level of tax doesn’t induce a very
large impact. (Elastic rm)
fi
ffi
fi
ff
ff
fi
fi
fi
fi
ff
fi
fi
ffi
ff
ff
fi
fi
ffi
ff
ffi
fi
fi
ffi
fi
ffi
ffi
ff
 Tax is an e cient way to change behaviour but very unpopular, from a political pov.
3rd solution: Emission Permits
Hybrid solution: mixture of regulations and taxation.
Government xes R: if you want to emit more you have to buy permits and than people can grade.
EU ETS= European Emission Trading System, which is market where each rm is endowed with a xed number
number of permits (es. In rm A, E=EA) based on historical data.
Some activities are not subject to thus type of regulation (at the moment 50%, goal of the EU is to enlarge it).
Permits can be traded opening a secondary market. Firm can move away from xed number assigned from
regulators, buying permits if they want to emit more than EA or selling it they emit less.

Price changes weekly (60€ for ton). Initially, price was lower and quantity larger, now they are less and expensive.
Price depends on demand and supply which in return depends on number of permits available.

Countries pay a marginal cost of reduction with the price of a permit.

For rm1, which faces higher marginal cost, is cheaper to buy permits than to reduce
the amount of reduction.
Each rm should compare its margins cost to what is available on the market.

First entitlements are provided and then they are allowed to trade to reach an e cient outcome. (Coasian solution)

Using a tax-> you get a revenue. The tax discourage an activity providing impediment. Tax revenue can be used to
make population less hostile through rebates and refunds.

COMPARISON OF TAXES AND REGULATION

If there are perfect info about marginal bene t or marginal cots: you can implement a fee or a regulation to obtain an
e cient level

Martin Weitzman: What happens if policy maker has imperfect information about cost?
=> There is uncertainty about the actual cost of reduction emission at the time tax/regulation will be enacted.
Supposing you have to enact tax/regulation that will last for a long time its but there is uncertainty about the future
regarding the cost while marginal bene t is known. When the uncertainty will be resolved, rms will know the cost.
policy maker is making decision in a condition of asymmetric information (very common in eld of climate change )
Marginal cost of reduction can be high, c’0 or low, c’1
Choice depends on the inclination of b’.
- b’ at- > tax
- B’ steep-> regulation

Impact on stock of emission accumulated over the past years large, the damage is still there. Applying a regulation
reducing emission has small impact, the bene t won't change that much. In this case the tax is the best tool:
economist argue that regulation have marginal of errors and cost of reducing emission in is uncertain
ffi
fl
fi
fi
ffi
fi
fi
fi
fi
fi
fi
ffi
fi
fi
fi
fi
while b’ s stable (es. Carbon tax).

1st case: Regulation

EU commission sets the amount of permits and reduction that should be carried out without knowing marginal cost.
Ideally, if it could see marginal cost, it could nd the e cient level, but in reality there is 50% probability of getting R’0
and 50% R’1. Commission must try to minimise the cost of mistake choosing a level of regulation in between (R̄).

In both cases there will be a deadweight loss:

- DWL0 if high marginal cost materialise (level of regulation is too high)
- DWL1 if low marginal cost materialise (level of regulation is too low).

-> Diagram with deadweight loss under regulation

2nd case: Taxes

In the case of rm 0, optimal fee is f0 and for rm 1 optimal fee is f1 v
EU emits an emission fee F̄ to minimise losses.
Firms reduce up to the point where cots of emission (reduction) equals cost of the emission fee.

• Firm 0 gets to R0 (actual reduction of firm0). Ideal fee would’ve been too
high,
thus firm 0 reduces less than eff. level.
• Firm gets to R1 (actual reduction of firm1). Ideal fee would’ve been too low,
thus firm 1 reduces more than eff. level.

Aim of the policy maker is to minimise the sum of the deadweight losses.
Bringing the DWL in the same diagram

=> What’s best between tax/ regulation depends on the slope of the b’.

1 case: Inelastic (b’ is flat) 2 case: elastic (b’ is steep)

DWL of regulation> DWL fees=> FEES DWL of regulation< DWL fees => REGULATION
fi
fi
fi
ffi
ETS covers 40% of the emission in the EU to put pressure on there rms to produce less emissions and reduce
number of permits making them more expensive. ETS includes maritime transportation and aviation. Permits are
given based on historical data.
“Grand fathering”: assigning the right to people who have been doing it already (countries who have been emitting a
lot receive a lot of permits). The word comes form CROW legislation in the US which kept black people out of
political system from voting by saying they could vote only if they had high education level. However f grand father
had the right to vote these people could vote.

Aviation is subjected to ETS.

Each time a plane ies, the company has to buy permits. If plane doesn’t y the permits that aren’t used are sold and
another company buy them to produce emission.
Ex. For aviation n 80% are free and 20% of permits are purchased to damage not too much the sector

Aim: carbon neutral by 2050

fl
fi
fl