资源简介

(共44张PPT)
20
20.1 Taxation and Economic Efficiency
20.2 Optimal Commodity Taxation
20.3 Optimal Income Taxes
20.4 Tax-Benefit Linkages and the Financing of Social Insurance Programs
20.5 Conclusion
Tax Inefficiencies and Their Implications for Optimal Taxation
20
Introduction
Arnold Harberger, one of the pioneers of the general equilibrium tax incidence model, wrote of his experience in Indonesia, where cars are taxed more heavily than motorcycles.
“Three-wheel cycles were converted, by artful additions, into virtual buses, or at least taxis. Sometimes a single bench was added, with the passenger looking backward. Other times the cycle was stretched at the back, with two benches going down each side, and maybe even with an extra little running board cutting laterally across the rear (where the rear bumper of the car would be). I must say I was truly astounded when I saw my first eight-passenger motorcycle.”
This example highlights a simple fact: markets do not take taxes lying down. If there is some action that market participants can undertake to minimize the burden of taxation, they will do so.
20
Taxation and Economic Efficiency
Usually, the market produces efficient outcomes.
Taxes interfere in the market and reduce efficiency.
People substitute away from the taxed product, using less-efficient alternatives.
Eight-person motorcycles in Indonesia
Some taxes have much larger efficiency costs than others.
B
C
Price per
gallon (P)
Quantity in
billions of
gallons (Q)
0
S1
D1
P1 = $1.50
Q1 = 100
S2
E
Tax =
$0.50
Q2 = 90
P2 = $1.80
P3 = $1.30
F
G
D
A
Taxation and Economic Efficiency: Graphical Approach
20.1
Deadweight loss, DWL
A tax is imposed, and supply shifts from S1 to S2. Deadweight loss occurs—triangle BAC.
20.1
Taxation and Economic Efficiency
Absent taxes:
price = social marginal benefit = social marginal cost
The tax drives a wedge between SMB and SMC, preventing mutually beneficial trades from occurring.
The units between 90 and 100 would have generated a consumer and producer surplus.
The foregone surplus from taxation is called the deadweight loss (DWL).
The size of the DWL depends on elasticities.
B
Price per
gallon (P)
Quantity in
billions of
gallons (Q)
0
(a) Inelastic demand
S1
D1
P1
Q1
(b) Elastic demand
S2
P2
Q2
Tax
A
B
C
DWL
Price per
gallon (P)
Quantity in
billions of
gallons (Q)
0
S1
D1
P1
Q1
S2
P2
Q2
Tax
A
DWL
C
Elasticities Determine Tax Inefficiency
20.1
The deadweight loss of a given tax is smaller when the demand curve is less elastic than when it is more elastic.
Deadweight loss is caused by individuals and firms making inefficient consumption and production choices in order to avoid taxation.
The inefficiency of any tax is determined by the extent to which consumers and producers change their behavior to avoid the tax.
The more elastic is demand or supply, the larger the DWL.
20.1
Elasticities Determine Tax Inefficiency
In 1696, English King William III needed money to finance an ongoing war with France.
To raise this revenue, the king imposed a new tax based on the number of windows in a home, which was an indicator of a home’s value: more valuable homes would have more windows, and therefore be taxed more highly.
The problem was that individuals could minimize their tax by boarding up or otherwise removing windows. This rendered the tax a less efficient measure of home value, and created a deadweight loss by making it less pleasant (and therefore of less value to residents) to live in their homes.
20.1
Empirical Evidence: The Window Tax
Oates and Schwab (2015) presented a straightforward test for whether this was the case: they looked at whether the number of windows responded to “notches” in the tax schedule.
From 1747–1757, there was no tax if the house had fewer than ten windows, a tax of 6 pence per window for 10–14 windows; a tax of 9 pence per window if there were 15–19 windows; and a tax of 1 shilling (12 pence) per window for 20 windows or more.
The marginal tax rate jumped enormously upon building the 10th, 15th, or 20th window.
20.1
Empirical Evidence: The Window Tax
There are obvious spikes at 9, 14, and 19 windows, exactly what we would predict from the effects of this tax.
20.1
Empirical Evidence: The Window Tax
One could argue that perhaps there was some British custom or some other reason for having 9, 14, or 19 windows in a house.
Oates and Schwab address this point by using the fact that in 1761, the government added a provision that houses with 8 or 9 windows would pay 1 shilling per window.
This means that after 1761 there was suddenly an incentive to keep the number of windows to 7 or fewer. This is exactly what happened.
20.1
Empirical Evidence: The Window Tax
In data from 1761–1765, the authors find that 27% of houses have 7 windows, while only 5% have 6 windows, and 3% have 8 windows. This is a nine-fold increase from the 3% of buildings with 7 windows from just a few years earlier. Clearly, this was a response to the tax.
20.1
Empirical Evidence: The Window Tax
Keynes: “The avoidance of taxes is the only pursuit that still carries any reward.” Some examples of avoidance:
The Papal States taxed salt heavily, so Tuscan bakers stopped using it. Even today, Tuscan bread is saltless.
In the early 1980s, Cyprus’s building tax applied to finished structures. Homeowners put steel bars jutting out from their roofs to avoid the tax.
Thailand taxes business signs on the outside, with higher taxes for English-only signs. So, many signs have a bit of Thai writing in the corner or are hung on curtains inside the shop.
20.1
APPLICATION: Tax Avoidance in Practice
The formula for DWL is
and are the elasticity of supply and demand is the tax rate, Q and P are the quantity and price.
DWL rises with the square of the tax, so marginal DWL rises with the tax rate.
Marginal deadweight loss: The increase in deadweight loss per unit increase in the tax.
20.1
Determinants of Deadweight Loss
B
C
Price
of gas
Quantity of gas
0
S1
D1
P1
Q1
E
Tax =
$0.10
Q2
P2
P3
D
DWL
Q3
Tax =
$0.10
S2
S3
A
Determinants of Deadweight Loss
20.1
The marginal deadweight loss rises disproportionately with the tax rate.
20.1
Since marginal DWL rises with the tax rate, pre-existing distortions affect the efficiency of a new tax.
Preexisting distortions: Market failures, such as externalities or imperfect competition, that are in place before any government intervention.
A Tax System’s Efficiency Is Affected by a Market’s Preexisting Distortions
20.1
A Tax System’s Efficiency Is Affected by a Market’s Preexisting Distortions
A tax in a market with no preexisting distortions creates a deadweight loss equal to triangle BAC. A positive externality in the market has created a deadweight loss triangle EDF; imposing a tax on this market results in a deadweight loss that is larger by the area of trapezoid GEFH.
20.1
Progressive Tax Systems Can Be Less Efficient
Because the DWL rises with , progressive tax systems can be less efficient than proportional ones.
Example:
Suppose there are two people, one with a wage of $10/hour and one with a wage of $20/hour.
For both, a 10% rise in wages leads them to supply 10% more labor (elasticity of labor supply = 1).
Elasticity of labor demand is also 1.
20.1
Progressive Tax Systems Can Be Less Efficient
No Tax Proportional Tax Progressive Tax
Tax rate below $10,000 0 20% 0%
Tax rate above $10,000 0 20% 60%
Low-wage hours 1,000 864 1,000
Low-wage DWL 0 $115.7 0
High-wage hours 1,000 864 837
High-wage DWL 0 $231.4 $566.8
Total DWL 0 $347.1 $566.8
The deadweight loss is larger for the higher-wage worker despite the same reduction in hours worked.
20.1
Why is the deadweight loss larger for the higher-wage worker despite the same reduction in hours worked
In a competitive labor market, wage equals the marginal product of labor, so the high-wage worker has a higher marginal product of labor.
Society loses twice as much when the high-wage worker reduces her hours than when the low-wage worker reduces her hours.
Progressive Tax Systems Can Be Less Efficient
20.1
Progressive Tax Systems Can Be Less Efficient: Graphical Approach
Initially the government imposes an equal tax on the low-wage worker and the high-wage worker, which results in deadweight losses of triangles BAC and EDF.
20.1
Progressive Tax Systems Can Be Less Efficient: Graphical Approach
When system is replaced with no tax on low-wage workers, there is no DWL for this worker, but the DWL for the high-wage worker increases by the trapezoid GEFI. This results in an increase in deadweight loss because additional taxes must be collected from high-wage workers in order to collect the same revenues as before.
20.1
Government efficiency in taxation over time is maximized by tax smoothing, by having a relatively constant tax rate over time rather than high taxes in some periods and low taxes in others.
High-then-low tax rates produce a larger DWL then steady medium tax rates.
Governments Should “Smooth” Tax Rates Over Time
20.1
Hausman (2000): The DWL from taxes on wireless communications equals 53% of revenue.
This figure is high for three reasons:
High price sensitivity (elastic demand).
Large pre-existing distortions from imperfect competition.
High taxes—25% in some states, and in those states, the DWL is even higher.
APPLICATION: The Deadweight Loss of Taxing Wireless Communications
20.2
Optimal commodity taxation: Choosing the tax rates across goods to minimize deadweight loss for a given government revenue requirement.
Ramsey Rule: To minimize the deadweight loss of a tax system while raising a fixed amount of revenue, taxes should be set across commodities so that the ratio of the marginal deadweight loss to marginal revenue raised is equal across commodities, that is, .
Value of additional government revenues: The value of having another dollar in the government’s hands relative to its next best use in the private sector.
Ramsey Taxation: The Theory of Optimal Commodity Taxation
20.2
Inverse Elasticity Rule
If supply is infinitely elastic, the Ramsey Rule becomes:
is the optimal tax, is the elasticity of demand, and is some constant.
Optimal taxation therefore balances two rules:
Elasticity rule: Lower taxes on goods with more elastic demand.
Broad base rule: Better to tax a wide variety of goods moderately than few goods heavily.
20.2
Imagine that the government had only two goods it could tax, cereal and caviar:
Elasticity of demand for caviar is much higher than that for cereal.
The inverse elasticity rule would suggest that the government tax cereal much more highly than caviar.
This means taxing more heavily the good consumed by poor people.
This might hurt vertical equality.
Equity Implications of the Ramsey Model
20.2
Commodities are taxed or subsidized throughout the developing world.
Deaton (1997) studied the demand for subsidized goods in Pakistan, looking at their elasticity, and the income of people who consume it.
APPLICATION: Price Reform in Pakistan
Good Subsidy Elasticity Consumed by
Wheat 40% 0.64 Poor
Rice 40% 2.08 All
Oil/Fat 5% 2.33 Poor
20.2
APPLICATION: Efficiency Consequences of Taxes and Subsidies in Pakistan: Wheat (Inelastic Demand)
In the market for wheat, demand is fairly inelastic and supply is subsidized, leading quantity to increase from Q1 to Q2 with a deadweight loss of BAC.
20.2
APPLICATION: Efficiency Consequences of Subsidies in Pakistan: Rice (Elastic Demand)
In the market for rice, demand is very elastic, so when supply is subsidized the quantity rises by much more (from Q1 to Q2), and the deadweight loss is larger (BAC).
20.2
APPLICATION: Efficiency Consequences of Taxation in Pakistan: Oils and Fats (Elastic Demand)
In the market for oils and fats, demand is also very elastic, so even the small tax leads to a large reduction in quantity from Q1 to Q2, with a deadweight loss of BAC.
20.3
Most tax revenue in the United States and other developed countries is from income taxes.
Optimal income taxation: Choosing the tax rates across income groups to maximize social welfare subject to a government revenue requirement.
Social welfare function guides the trade-off between progressivity and efficiency.
Optimal Income Taxes
20.3
1. Everyone in society has the same utility functions:
…
2. Diminishing MU of income.
3. Total income in society is fixed, not determined by individual choices that might respond to tax rates.
4. Society has an equally-weighted utilitarian social welfare function:
A Simple Example
A Simple Example
20.3
Under these assumptions:
The optimal income tax system is one that leaves everyone with the same level of post-tax income.
People with income below average would receive a transfer to increase their incomes to average.
The marginal tax rate under this system is 100%.
If income responds to taxes, the optimal marginal tax rate is lower.
20.3
General Model with Behavioral Effects
With behavioral effects, taxes reduce hours worked.
At high tax rates, tax revenue falls with the tax rate; no one works under a 100% tax rate.
The optimal tax system trades off the efficiency cost of taxation against the benefits of redistribution.
The rule is to set the income for group such that
is MU for group , the marginal revenue, and is the value of government revenue.
20.3
The Laffer Curve
As tax rates rise from 0 to τ*, tax revenues rise; but when tax rates rise above τ* toward 100%, tax revenues fall.
20.3
General Model with Behavioral Effects
Optimal income taxation balances:
Vertical equity: Social welfare is maximized when those who have a high level of consumption, and thus a low marginal utility, are taxed more heavily, and those who have a low level of consumption, and thus a high marginal utility, are taxed less heavily.
Behavioral responses: As taxes rise on any one group, individuals in that group may respond by earning less income.
20.3
An Example: Optimal Income Taxes with Two Types
The ratio of marginal utility to marginal revenue rises as tax rates rise for any taxpayer, but this ratio for Mr. Rich is everywhere below the ratio for Ms. Poor. Optimal income tax rates are those that equate this ratio across taxpayers. Here, the optimal rates are 10% for Ms. Poor and 20% for Mr. Rich.
So far we have ignored tax-benefit linkages.
Tax-benefit linkages: Direct ties between taxes paid and benefits received.
Introducing these linkages enriches changes the story, since many payroll taxes are directly linked to benefits.
20.4
Tax-Benefit Linkages and the Financing of Social Insurance Programs
20.4
Tax-Benefit Linkages: Graphical Representation
A pure tax on labor would shift the demand curve from D1 to D2 and creating the deadweight loss triangle CAB. If those taxes are tied to benefits provided to workers, then supply shifts out to S2. Labor supply falls only to L3, and the deadweight loss triangle shrinks to GAF.
20.4
Perfect Linkage Eliminates the DWL
When workers value the tax-financed benefit so highly that they are willing to accept its full cost in lower wages, there is no change in employment when the tax is imposed.
20.4
Why doesn’t the private sector provide these benefits
Market failures may plague the market.
When are there tax-benefit linkages
The tax-benefit linkage is strongest when taxes paid are linked directly to a benefit for workers.
What is the evidence on tax-benefit linkages
Financing is borne by workers in the form of lower wages and not lower employment.
Issues Raised by Tax-Benefit Linkage Analysis
In the mid-1970s, states began to mandate insurance benefits for pregnant women.
These laws raise the cost of insuring, and hence employing, certain groups.
Compared to other groups in the same states, or the same groups in states without mandated benefits:
Wages fell.
Labor supply did not.
Suggests that benefit linkage is near complete.
20.4
EVIDENCE: A Group-Specific Employer Mandate
20.5
The fundamental issue in designing tax policy is the equity-efficiency trade-off.
Tax efficiency comes down to two key principles:
The more elastically supplied or demanded the good, the larger the deadweight loss from the tax.
The higher the tax rate, the larger the incremental deadweight loss of taxation.
Conclusion

展开更多......

收起↑