### Economics 157b Economic History, Policy, and Theory Short

```Economics 331b
Finale on economics
of energy regulation
1
We are heading into a major period of energy/climate-change
regulations. Here are some of the major economic issues:
1. Rebound effect
•
•
Energy efficiency standards affect the energy-intensity of new
capital goods
Because they lower the MC, they may increase utilization,
leading to the rebound effect.
•
•
Increase oil use leads to higher oil world price
This leads to higher total imports costs and macro
disturbance
3. Public finance issues
•
•
•
Regulation and energy taxes lead to higher prices
These lead to dead-weight loss when P > MC
This leads to “double dividend hypothesis” and to concern
about using standards (with no revenues) instead of taxes
(with revenues that can lower other taxes)
4. Cost of capital/discounting (later on this one)
2
Price of vmt
Economics of rebound effect
Effect of efficiency
improvement
“Rebound effect”
Before mpg
improvement
After mpg
improvement
G
3
3
Gasoline consumption
Economics of rebound effect
Assume that regulation increases energy efficiency of a capital good
from mpg0 to mpg1 . The question is whether the lower cost of a vmt
(vehicle-mile traveled) would offset the lower cost.
(1) vmt = f vmt (p vmt , p cars ), vehicle miles traveled,
(2) cars = f cars (p vmt , pcars ), number vehicles
From this we can get the following:
(3) Gasoline use = G = vmt/mpg
(4) p vmt = p gasoline /mpg
(5) d ln G / d ln mpg  -1   d ln vmt / d ln p vmt    d ln p vmt / d ln mpg 
But we know from (4) that  d ln p vmt / d ln mpg     d ln p vmt / d ln pgasoline  , so
(6) d ln G / d ln mpg  -1   d ln vmt / d ln p vmt    d ln p vmt / d ln p gasoline 
which gives us the important result:
(7) d ln G / d ln mpg  1   d ln vmt / d ln p gasoline 
So rebound effect is equal to the elasticity of vmt with respect to gasoline prices,
which we have observed in countless studies.
4
Empirical estimates of rebound effect
Basic results from many demand studies:*
Short-run gasoline price-elasticity on vmt = -0.10 (+0.06)
Long-run gasoline price-elasticity on vmt = -0.29 (+0.29)
Therefore, the rebound would be 10 to 29 percent of mpg improvement.
This can be applied to other areas as well.
Reference: Phil Goodwin, Joyce Dargay And Mark Hanly, “Elasticities of Road Traffic and Fuel
Consumption with Respect to Price and Income: A Review,” Transport Reviews, Vol. 24, No. 3,
275–292, May 2004, available at http://www2.cege.ucl.ac.uk/cts/tsu/papers/transprev243.pdf
5
Source: UK Energy Research Centre, The Rebound Effect
6
The “oil premium” refers to the excess of the social marginal cost of oil
consumption over the private marginal cost.
Analytically, this is
 "Oil

 supply   monopsony 
 macro




 price   premium

 externality 



 



 supply   oil 

- 


 price   taxes 
 monopsony 
 macro
  oil 
 

-




 externality   taxes 
7
Basic argument. The point is that the US has market power in the world
oil market. By levying tariffs, we can change the terms of trade (oil
prices) in our favor.
Regulation and taxes are a substitute for the optimum tariff.
Example:
• world supply curve to US: Q = Bpλ , λ>0
• US cost of imported oil = V = pQ = B-1Q(1+1/ λ) , k an irrelevant constant
• marginal cost of imported oil = V’(Q) = (1+1/λ) B-1Q1/ λ = p (1+1/ λ)
So optimal tariff is ad valorem:
τ = 1/ λ = inverse elasticity of supply of imports
Reference: D. R. Bohi and W. D. Montgomery, “Social Cost of Imported Oil and UU Import
Policy,” Annual Review of Energy, 1982, 7, 37-60.
8
Basics of deriving oil (monopsony) premium
Here is a more rigorous proof of the oil-import premium:
(1) Domestic or consumer prices = v = import price + tariff = p + 
(2) Oil supply to the US: Q  Bp   B(v -  ) , or p = (Q/B)1/
(3) Assume that marginal value of oil in US is constant at v
(infinitely elastic demand)
So we want to
(4) max vQ  pQ  vQ  bQ[1 1/ ]
Q
Maximizing with respect to Q leads to a maximum value of tariff:
(5) v  p    [1  1 /  ] p
Since consumer prices are equal to marginal value of v ,
we set the optimal tariff as
(6)  = (p  v) / v  1 / 
= inverse elasticity of import supply.
Notes: (1) This does not have to be a tariff. It is really a “shadow
price” on oil imports. (2) Example of “Ramsey tax theory.”
9
P, MC of oil
MSC
S
Import
free-market
imports
“Optimal
Tariff” at
Optimized oil
imports
D
10
Q(“optimal”)
Q(free market)
Imported oil
10
Macroeconomic externality
Somewhat more tenuous is the macroeconomic externality.
Idea is that there are impacts of changes in oil prices on macro
economy because of inflexible wages and prices.
So have another linkage:
(oil  price)
d(realGDP)
(realGDP)


d(oil  consumption) (oil  price) (oil  consumption)
The second term was discussed in optimal tariff. The first term
comes from macroeconomics (see next slide).
This, however, is very controversial and the estimates are not
robust.
11
Macroeconomic externality
A standard macro/oil-price equation with “good” results.
Dependent Variable: LOG(GDPQ_BEA)
Method: Least Squares
Sample: 1971Q1 2009Q1
Included observations: 153
Variable
C
LOG(RPOIL08)
TIME
R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
Coefficient
Std. Error
t-Statistic
Prob.
4.399688
-0.017205
0.007508
0.027266
0.003546
3.89E-05
161.3626
-4.851700
193.0151
0.0000
0.0000
0.0000
0.996011
0.995957
0.021228
0.067591
373.8431
18724.88
0.000000
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
8.841860
0.333864
-4.847622
-4.788202
-4.823485
0.166943
12
Effect of oil prices on real GDP
(log = fractional; multiply by 100
to get percent difference)
-.035
-.040
-.045
-.050
-.055
-.060
-.065
1970
1975
1980
1985
1990
1995
2000
2005
13
Macroeconomic externality
Simplified derivation:
( oil  price )
d(realGDP )
(realGDP )


d( oil  consumption) ( oil  price ) ( oil  consumption)
[ln(realGDP )] realGDP
oil  price



Q
[ln( oil  price)] oil  price
  15000
 macro 
 premium  
7


Assume that  = 2 and  = -.014
.017  15000
 macro 

 \$18 / bbl
72


We can also derive that monopsony/macro = ε[GDP/pQ]
14
Updated estimates
Cited in Hillard G. Huntington, The Oil Security Problem, EMF OP 62,
February 2008.
15
Numerical example for US
Variable
Elasticity of supply of oil
0.1
1
Domestic production (10^6 bbls/day)
6
Imports (10^6 bbls/day)
14
Domestic demand (10^6 bbls/day)
20
Global production (10^6 bbls/day)
85
Oil price (2009 \$/bbl)
50
Elasticity of demand for oil
Elasticity of imports w.r.t. oil price
Optimal tariff on oil (\$/bbl)
Optimal tariff on oil (c/gallon)
5
-0.5
2.9
\$ 17.33
41
8.0
\$
6.28
15
30.5
\$
1.64
4
16
Optimal tariff argument on oil taxes
τ = 1/ λ = inverse supply elasticity.
Complications: Formula actually is

 S , ROW S ROW   D, ROW DROW
S ROW  DROW

1  79  (.5)  65
 8.0
79  65
Some notes:
1. Supply elasticity depends critically on whether oil market is at full
capacity (2007 v. 2009). Very inelastic in full capacity short run; quite
elastic when OPEC adjusts supply. (See next slide.)
2. The optimal tariff in \$ terms depends upon the initial price because it is
an ad valorem tariff.
3. The externality is a global externality for consuming countries because
it is a globalized market.
4. Note this is a pecuniary, not a technological externality. So it is a zerosum (or slightly negative-sum) game for the world. This has serious
strategic implications and suggest that the diplomacy of the oil-price
externality is completely different from true global public goods like
global warming.
17
Price
Short-run
production
capacity
Production
18
The “double dividend” hypothesis
Some have argued that using “ecological” or environmental
taxes has a double dividend:
1. Get environmental benefit when P < MSB.
2. Can use revenues from ecological taxes to reduce other
burdensome taxes
In economics, the burden is measured as “deadweight loss”
(DWL)
[Related issue in current context is whether the additional debt
incurred by the stimulus package has such high DWL that
the net economic effect is negative (Kevin Murphy).]
19
The dead weight loss of taxes/regulation
P, MC of oil
If add new taxes (regulation):
Additional revenues = D – B
Additional DWL = C + B
MDWL/Mtaxes =(C+B)/D-B)
≈ (B)/D-B)
S +T1+ T2
[When are you on the wrong side
of the peak of the Laffer curve?]
P2
D
C
S + T1
P1
E
B
A
S
P0
Imported oil
20
20
Thoughts on double dividend hypothesis
1.
Do not have DWL if raising price to the Pigovian level; actually
lowing the DWL from a “subsidy”
-
Actually a little more complicated because need to look at
existing taxes (e.g., gasoline) and complementarity/substitution
patterns with other goods and services.
2. A regulation is a tax with the revenues rebated to the polluter. If use
regulations rather than taxes, you therefore lose the second half of
the double dividend.
-
This is key argument in cap and trade v. carbon taxes (or more
generally regulation v. taxes)
3. If standards are beyond Pigovian levels, then can incur serious DWL
if do not attend to the revenue side of the issue.
4. Empirical estimates of MDWL of taxes: all over the place from \$0.2 to
\$2 per dollar of revenue.
21
```