Ownership Change, Incentives and Plant Efficiency: The Divestiture of U.S. Electric Generation Plants James Bushnell UC Energy Institute Catherine Wolfram.

Ownership Change, Incentives and Plant Efficiency:

The Divestiture of U.S. Electric Generation Plants

James BushnellUC Energy Institute

Catherine WolframUC Berkeley

Tenth Annual POWER Research ConferenceMarch 18, 2005 Berkeley, CA

Electric generating plant divestitures

• The ownership of electric generating plants in the U.S. changed dramatically between 1998 and 2001.

- Over 300 plants, representing nearly 20% of U.S. installed generating capacity, changed ownership.

• Divestitures were encouraged by state regulators as part of the industry restructuring.

- Easy measure of stranded costs.

- Addressed concerns about the ability of a firm that was vertically integrated into transmission to exercise market power.

• Old owners (investor owned utilities) faced cost-of-service regulation; new owners face market incentives.

Electricity industry restructuring and economic efficiency

• There is heated debate in academic and policy circles about how to design restructured markets.

• All of this should be moot if restructuring doesn’t improve economic efficiency.

• Several academic and policy studies have examined the effects of restructuring on cost efficiencies:

- Top down: DOE (2003), SEARUC (2003), CAEM (2004).

- Bottom up: Markiewicz, Rose and Wolfram (2004) looks at changes in operating efficiency at IOU plants facing new incentives; Kleit and Reitzes (2005) looks at dispatch efficiency.

Measuring the effects of divestitures on plant efficiency

• Ideally, we would like to evaluate changes in each plant’s total factor productivity after divestiture.

• Unfortunately, data on employment, maintenance and capital inputs is not publicly available after divestiture.

- We’re exploring data availability from various sources.

• For now, we’re analyzing whether fuel efficiency changed at plants that were divested.

- Rich data are available from the Environmental Protection Agency.

- Implicit assumption that production is Leontief in fuel and other inputs.

There may be no effect of divestiture on fuel efficiency

- A generating unit’s fuel efficiency is an immutable technological characteristic.

- IOUs may have made poor investment decisions, but at least they knew how to operate their own plants.

…or there may be either positive or negative net effects.

• Heat rates could go up following divestiture.

- It will take the new owners some time to master the idiosyncrasies of the plant.

- New owners may cycle the plants more (e.g. to exercise market power) at the cost of higher heat rates.

• Heat rates could go down following divestiture.

- IOUs had fuel adjustment clauses, so they didn’t do everything possible to minimize heat rates.

- New owners might be able to renegotiate union contracts so they can promote good operators.

Data set

• We are analyzing electric generating units that are:

- Fossil-fuel fired (i.e. not nuclear or hydroelectric).

- Included in the EPA Continuous Emissions Monitoring System data base.

- In regions where divestitures were prominent (this excludes, for instance, all the plants in Florida).

• We have hourly fuel input and output data on ~1000 generating units for 7 years (1997-2003).

We analyze plants in the shaded states.

Empirical specification

• To analyze divestitures, we estimate versions of the following equation:

• for unit i in time period t, where

- HR is the unit’s heat rate (inverse of fuel efficiency),

- t is either monthly or hourly,

- Divestiture = 1 after divestiture (if applicable),

t are month dummies, θi are unit dummies, and

- we specify the relationship between HR and Gross Load with care.

ituniti

monthtititit GrossLoadeDivestiturfHR ,,),ln(,)ln(

Example of what we’re labeling a Divestiture effect.

Unit

Year

Divested Not divested

1997 14 11

2003 12 10

Average Heat Rates for 2 units across 2 years

Example of what we’re labeling a Divestiture effect.

Unit

Year

Divested Not divested

1997 14 11

2003 12 10

Our Divestiture effect:

(12-14) – (10 -11) = -1

Average Heat Rates for 2 units across 2 years

Our estimates control for:

- differences in average heat rate levels across

divested and not-divested units.

- general trends in heat rates.

Example of what we’re labeling a Divestiture effect.

Unit

Year

Divested Not divested

1997 14 11

2003 12 10

Our Divestiture effect:

(12-14) – (10 -11) = -1

Average Heat Rates for 2 units across 2 years

Our estimates control for:

- differences in average heat rate levels across

divested and not-divested units.

- general trends in heat rates.

Our Divestiture effect in logs:

(ln(12)-ln(14)) –

(ln(10) –ln(11)) -.05

The basic Divestiture effect: 2% reduction in Heat Rate

 Dependent Variable: ln(Heat Rate)

OLS IV

Only Divested

UnitsWith

Controls

Only Divested

UnitsWith

Controls

Divestiture -0.022** -0.021** -0.022** -0.013

(0.010) (0.010) (0.009) (0.009)

ln(Q) -0.405*** -0.364*** -0.380*** -0.106

(0.024) (0.018) (0.068) (0.120)

Month-Year Fixed Effects No Yes No Yes

Unit or Unit-Interval Fixed Effects Unit Unit Unit Unit

R2 .56 .54

N 22,506 75,066 19,501 64,882

See Table 4 in the paper.

Why do we see lower heat rates after divestitures?

• Utilities only sold plants that were ripe for improvements.

In states with major divestitures, IOUs sold nearly all their thermal capacity.

StateTotal Fossil

Fuel MW IOU % in 2003

CA 36600 2%

CT 5750 0%

DC 590 0%

DE 3084 0%

IL 32486 5%

MA 11983 0%

MD 10373 1%

ME 2860 0%

MT 2448 6%

NJ 15258 4%

NY 28196 19%

PA 33424 5%

StateTotal Fossil

Fuel MW IOU % in 2003

IN 26342 71%

KY 19482 49%

NH 2523 43%

OH 32782 55%

VA 16501 63%

Why do we see lower heat rates after divestitures?

• Utilities only sold plants that were ripe for improvements.

Why do we see lower heat rates after divestitures?

• Utilities only sold plants that were ripe for improvements.

• Utilities deferred maintenance before they sold the plants; merchant firms have been regaining lost ground.

Why do we see lower heat rates after divestitures?

• Utilities only sold plants that were ripe for improvements.

• Utilities deferred maintenance before they sold the plants; merchant firms have been regaining lost ground.

- But we find no evidence of heat rates deteriorating before the divestitures (see Table 7).

Why do we see lower heat rates after divestitures?

• Utilities only sold plants that were ripe for improvements.

• Utilities deferred maintenance before they sold the plants; merchant firms have been regaining lost ground.

• Merchant firms faced new incentives—incentive effect.

• New owners shook things up—ownership effect.

Table 7: evidence on incentive and ownership effects

IOU plants showed improvements in states with rate freezes,

suggesting incentives matter.

Evidence on effects of ownership transfers less clear.

Dependent Variable: ln(Heat Rate)

OLS

All Units

All Units

Divestiture -0.024** (0.011) External Only -0.003 (0.013) External Both -0.023 (0.018) Internal Subsidiary -0.067*** (0.024) Incentive Regulation -0.019** -0.019* (0.010) (0.010) ln(Q) -0.364*** -0.364*** (0.018) (0.018) Month-Year Fixed Effects Yes Yes R2 0.54 0.54 N 75066 75066

Why do we see lower heat rates after divestitures?

• Utilities only sold plants that were ripe for improvements.

• Utilities deferred maintenance before they sold the plants; merchant firms are regaining lost ground.

• Merchant firms faced new incentives—incentive effect.

• New owners shook things up—ownership effect.

What’s being done differently at divested plants?

• Are merchant firms operating at the “sweet spot” more?

• Are merchant firms changing other aspects of operations (combustion optimization software, fuel switching, etc.)?

Unit heat rate profiles

5

7

9

11

13

15

17

19

21

23

25

0 2 4 6 8 10 12 14 16 18 20 22

Capacity Interval

Ave

rag

e H

eat

Rat

e (m

mB

tu/M

Wh

)Overall < 100 MW > 100 MW

Implied magnitude

• Our estimates suggest that plants have ~2% lower heat rates after the divestiture.

• At current fuel prices, this amounts to about $0.6/MWh.

- At the plants that were divested, this adds up to savings of nearly $500 million per year.

- Scaling this up to all thermal plants nationwide, this could add up to almost $1.5 billion per year.

• Improving fuel efficiency helps achieve environmental goals,

especially with respect to CO2

Conclusions

• In light of the California electricity crisis, it is useful to remind ourselves about the potential efficiency gains of restructuring.

• This paper focuses on fuel efficiency at existing generating plants.

• Evidence is positive—owners respond to incentives.

• Additional efficiency gains possible through:

- More efficient long-term (capital) investment.

- Incentive regulation for transmission and distribution.

- Reallocating output across plants.