california’s electricity crisis what

8/2/2019 Californias Electricity Crisis what

1/35

The California electricity crisis threatens notonly the economic well-being of ratepayers inCalifornia but the economic well-being of theUnited States as well. Unfortunately, the polit icaland economic commentary surrounding the cri-sis is shedding more heat than light.

California has been victimized by severalsimultaneous and severe supply and demandshocksmost notably, a run-up in wholesale nat-

ural gas pricesthat are outside the states poli t-ical control. Those shocks were made moresevere by air pollution regulations and retailprice controls. Although Californias deregula-tion of the electric utility business is beingblamed for the crisis by both the political left andthe political right, we find that the 1996 restruc-turing law has little to do with the run-up inwholesale power prices. That law is primarilyresponsible for the blackouts, however, in that itprohibits utilities from passing on increases infuel costs to consumers.

Virtually all the increase in wholesale pricescan be explained by increases in production costsand overall scarcity. While there is some evidenceof the existence of excessive generator market

power (created not by the unfettered exercise offree markets but by poorly conceived regulation),it is relatively minor and responsible for only asmall fraction of the price spike, if it exists at all .

We find little evidence to support the argu-ment that environmentalists are primarily toblame for the crisis. We likewise are unconvincedthat, had the state allowed utilities to enter intolong-term contracts with generators, the crisis

could have been either averted or made less severe.None of the remedies thus far proposed

such as a state takeover of the industry, the so-farminimal increase in power rates, energy conser-vation subsidies, prohibitions of wasteful ener-gy use, more vigorous wholesale price controls,or the adoption of long-term power contractswith generatorswill get the state through thenext two years without frequent and widespreadblackouts and significant economic damage. Infact, all of those alleged remedies would makematters worse.

The only remedy to the crisis is the elimina-tion of the retail rate cap and the institution ofreal-time pricing mechanisms for the largest seg-ment of demand possible.

Californias Electricity CrisisWhats Going On, Whos to Blame, and What to Do

by Jerry Taylor and Peter VanDoren

_____________________________________________________________________________________________________

Jerry Taylor isdirector of natural resource studiesat the Cato Institute. Peter VanDoren iseditor of CatosRegulation magazine.

Execut ive Summary

No. 406 July 3, 2001


2/35

Introduction

Skyrocketing wholesale power prices inCalifornia and the daily threat of brownoutsand blackouts have cast a pall over the meritsof electricity deregulation. Liberals, led byCalifornias governor, Gray Davis, blame therestructuring law passed in 1996 for the cri-sis, arguing that it left the state vulnerable tomarket manipulation by greedy power pro-ducers. According to Davis, the crisis is large-ly artificial but nonetheless a harbinger ofthings to come, not only in national electric-ity markets but in industries throughout theeconomy, if we continue our mad rushtoward laissez faire.

Conservatives for the most part agree thatthe 1996 reforms are primarily responsiblefor the crisis. They charge, however, that theregulations attached to those reformspri-marily the prohibition of long-term con-tracts between utilities and power generatorsand the imposition of a centralized daily spotmarketare largely responsible for the pricespike. The political right argues thatCalifornias regulations, crafted by environ-mental activists and anti-growth consumergroups, have long discouraged investment in

new generating capacity and that the black-outs are a long-overdue flock of chickenscoming home to roost.

While both sides are busily settling politi-cal scores, the real story of what happened inCalifornia is largely absent from most analy-ses. Accordingly, the important lessons thatthis crisis teaches about regulation and elec-tricity are largely being overlooked: retailprice controls are a recipe for disaster, andstate regulators have little idea how to orga-nize markets and certainly shouldnt attempt

to do so under the mantle of deregulation.

The Anatomy ofCalifornias Deregulation

On the last day of the 1996 legislative ses-sion, Californias legislature adopted by

unanimous vote A.B. 1890, the first of manystate attempts to deregulate the electricitybusiness. The roots of the confusion overwhat California didand did not deregulatebegan with the PR campaign to sell the billUnlike deregulation of the airline and truck-ing industrieswhich largely curtailed regu-latory oversight of those industriesthederegulation of the electricity industry inCalifornia was heavily prescriptive and noton balance, a loosening of regulatory over-sight at all. A.B. 1890 simply replaced the oldset of regulations with a new set of regula-tions, some of which were less interventionistthan the old, some of which were more.

The central thrust of A.B. 1890 was to cre-ate a competitive retail market for electricity

Tradit ional rate-of-return regulation of state-sanctioned monopoly power companies hadnot kept electricity prices at reasonable levelsBy the mid-1990s, Cali fornia electricity priceswere 35 percent higher than the U.S average,and California residential customers paid 35percent more than the average U.S. residen-tial customer.1 Those excessive costs threat-ened to slow expansion in California andmake the grid itself obsolete as ratepayersfled to nonutility power providers.2

The remedy was to allow competition

among and free entry of electric generators.The generators owned by incumbent utilitieswould compete with nonutil ity power gener-ators for business, and customers couldchoose from whom they wished to purchasepower. Competition would protect con-sumers from the excessive investments andcost overruns that occurred under tradition-al cost-of-service regulation.

The Crusade against M arket PowerWhile virtually all the nonutility interest

groups endorsed this reform in principlemany of thempart icularly the large indus-trial users of electricity, the independentpower producers, and the consumer rightslobbyfeared that incumbent utilities woulduse their control of the electricity grid (theincumbent-owned network of wood, steeland wire that connects power plants to con-

2

Whi le both sidesare busi ly set tl ing

pol i t i cal scores,the real story of

what happened i nCali for ni a is

largely absentf rom most

analyses.


3/35

sumers) to disadvantage independent gener-ators. Incumbent utilities would excludeindependent generators from the grid, priceaccess exorbitantly, or load the grid with theirown power and thus use congestion torestrict entry. Or utilities could charge retailcustomers below-cost rates while recoupingthose losses through excessively high trans-mission and distribution rates for competi-tors. Because most analysts assume that thegrid is a natural monopoly (building alterna-tive grids is widely thought to be prohibitive-ly costly and impossible politically because offierce community resistance), it was fearedthat allowing nonutility firms to enter thegeneration market would not result in a com-petit ive market for wholesale power.

Reformers thus concluded that electricityderegulation would require a whole new setof regulations and government interventionsto ensure that a competitive market wouldarise. The important provisions of A.B. 1890follow.3

Mandatory Open Access: Utility compa-nies must allow any generator access tothe electricity transmission systemunder terms, conditions, and pricesestablished by the state.

Vertical Disintegration: Incumbent utili-ties are to become transmission anddistribution companies, divestingthemselves of generators. And thedivested generators can sell power onlyto a state-managed power exchange.

Centralized Power Exchange:Any electric-ity the incumbent utilities need fortheir default customers (those who donot switch to competitive suppliersunder the retail choice program) has tobe purchased from a centralized, state-

managed power exchange. Indepen-dent marketers can buy through theexchange voluntarily. The exchangecreates an electricity supply curve fromgenerators hourly willingness-to-pro-duce offers. The hourly price is set bythe highest-cost producer whose out-put is necessary to meet estimated

demand from the utilities. Utility Retail Price Caps:For those cus-

tomers who remain with the threeincumbent utilities (Pacific Gas andElectric, Southern California Edison,and San Diego Gas and Electric), retailrates are frozen at levels 10 percentbelow June 1996 rates until the incum-bent utilities recover their strandedcosts or March 2002, whichever occursfirst.

Independent System Operator (ISO):Theday-to-day operation of the grid (thatis, the management of electricity trafficalong the wires) is directed, not by theutilities that actually own the grid, butby a nonprofit organization governed

by a 26-member advisory board of rep-resentatives of grid users. The ISO isfurther empowered to procure electrici-ty on an emergency basis if on any givenday the amount of power procured inthe centralized power exchange isinsufficient to meet demand.

Stranded Costs and Retail PriceControls

Although utilities were never particularly

excited about the proposed change from tra-ditional rate-of-return monopoly regulationto managed competition, they were particu-larly worried that, were they forced to com-pete with independent power generators inthis new wholesale marketplace, some oftheir generating assets (nuclear power plantsand long-term contracts with nonutilitypower producers) would not produce suffi-cient revenues to recover their total costs.This would reduce the market value of thoseassets to much less than their book value

on the utilities balance sheets. Electricityanalysts coined the phrase stranded coststo describe the difference between book andmarket values of certain generating assets. InCalifornia, stranded costs were estimated tobe between $21 billion and $25 billion.4

Estimates for the United States ranged from$70 bill ion to $200 billion; the higher f igure

3

Reformers con-cluded that elec-tr i ci ty deregulat ion wouldrequi re a wholenew set ofregulati ons andgovernmentinterventions.


4/35

was double the total shareholders equity inthe industry.5

Util ities and several prominent academicsjustified recovery of those stranded costs byarguing that a social contract existedbetween investors in utilities and state regu-lators. In their view, the investors and thegovernment agreed that, if the companieswould serve all customers in an area as astate-sanctioned monopoly, the investorswould receive a regulated but reasonablereturn on their investment.6 Thus, deregula-tion without compensation of investors forthe capital losses they suffer as a resultbreaches the contract. It takes investorsproperty without compensationa violationof the Fifth Amendments Takings Clause

and raises the cost of capital in any other sec-tor of the economy in which potentialchanges in public policy create risk.

Economists disagree about whetherstranded-cost recovery is either efficient orfair.7 Those who oppose stranded-costrecovery argue that investors surely knowthat the rules of the regulatory game are sub-ject to change, and investors can and do takethat into account when investing theirmoney. You place your investment bets andyou take your chances.8

Moreover, if taxpayers were constantlyreimbursing businesspeople for investmentsthat become less attractive once the rules ofthe game are changed, beneficial regulatoryreforms might never occur. Furthermore,many of those stranded costs were enthusias-tically embraced by the utilities at the time.Few utilities were dragged kicking and scream-ing into those costly investments; many util ityexecutives wisely passed them by.

Nevertheless, A.B. 1890 allows recovery ofutilities stranded costs through the imposi-

tion of a mandatory competitive transitioncharge on all ratepayers. Of course, it wouldbe hard to sell ratepayers on a deregulationscheme that raised rates in the short run viasome incomprehensible competitive transi-tion charge. So A.B. 1890 imposes a ratefreeze on industrial and large commercialutility customers and a 10 percent reduction

in residential and small commercial utilityrates. Those rate controls are to stay in placeuntil a utility recoups its allotted share ofstranded costs, or no later than March 2002.

How could the state mandate a rate cutand impose a competitive transition chargeat the same time? Simple: it issued tax-exemptbonds through the off-budget CaliforniaInfrastructure and Economic DevelopmentBank to amortize the rate cut over a period oftime longer than the period during which therate cut was actually given to consumers. SoCalifornia and federal taxpayers pay for therate relief in the form of tax receipts forgonebecause the interest is tax-exempt.

The Mirage of Retail Choice

Under A.B. 1890, consumers not onlyenjoy a rate cut, they also are allowed tochoose their power company in much theway they choose their long-distance tele-phone company. And the competitionamong electric generators is supposed tokeep rates down. But A.B. 1890 sets the ratefor default electric service from the incum-bent utilities for the period 1998 to March2002 at 10 percent less than the 1996 rate;the difference is to be paid for, in part, by thetax-exempt bonds. New independent firms

have not been able to underprice this defaultservice. Only 1 percent of Californias resi-dential ratepayers had switched suppliers byearly 2000 despite large advertising invest-ments by power marketers.9 One of the lead-ing power marketers, Enron, actually aban-doned the California residential market afteronly three weeks because consumers had lit-tle incentive to leave their traditional suppli-ers. Competition was largely confined to thecentralized power exchange, where nonutilityfirms competed in a mandatory day-before

spot market to sell powernot directly toconsumers but to the incumbent util it ies act-ing as default suppliers to customers.

The Precrisis MarketAlthough the public may not have

grasped exactly what A.B. 1890 was all aboutbeyond some vague notion that competi-

4

Cal if orni a andfederal taxpayers

pay for the ratereli ef i n the f orm

of tax receipts for -gone because the

int erest i s tax-exempt.


5/35

tion was on the way because of deregulation,energy analysts were less certain that any ofthis would work. While most analysts feltthat, on balance, the new regime would provea valuable first step toward further deregula-tion, some, including the authors of thispaper, seriously doubted it and wrote asmuch at the time.1 0

Still, Californias electricity market underthe new regime appeared to work reasonablywell from April 1998 through the spring of2000. Even though retail competition forelectricity ratepayers never fully materializedbecause of the price controls on default ser-vice, wholesale electricity prices averaged $30per megawatt-hour (MWh), or 3 cents perkilowatt-hour (kWh), in 1998 and 1999.11

Those low prices allowed utilities to earn areturn on sales and recover stranded costseven with the retail rate cap.12 Between April1998 and April 2000, when the wholesalecost was lower than the retail price cap, thethree incumbent utilities retired $17 billionin debt.13 Incumbent utilities were also ableto recoup some stranded costs by selling fos-sil-fuel power generators with a total bookvalue of $1.8 billion for a combined $3.1 bil-lion retail price.1 4

Analysts did not predict that the crisis

would occur.1 5

University of California-Berkeley, economist Severin Borenstein, forexample, remarked: If we [economists] hadunderstood better, we would have warned bet-ter. . . . There were many things we didnt see.16

In its summer 2000 system reliability report,the North American Electricity ReliabilityCouncil estimated the reserve margin in theCalifornia-Mexico interconnection to be 15percent over the estimated July peak demand, alarger reserve margin than in several otherregions of the United States.17 The assessment

did warn that the region may not have ade-quate resources to accommodate a widespreadsevere heat wave or higher than normal genera-tor forced outages,18 but NERC was muchmore worried about the supply-demand imbal-ance in New York and New England than inCalifornia.19 Even in March 2000 the CaliforniaEnergy Commission forecast a weighted-aver-

age price for 2000 of 2.85 cents per kWh and apeak monthly weighted-average price inSeptember of 4.5 cents per kWh. The actualweighted-average prices were 11.3 cents perkWh for 2000, and the highest monthly averageprice was 37.2 cents per kWh in December.20

And as late as November 2000 NERC argued inits 200001 winter assessment that generationin all regions would be able to meet demandshould normal weather prevail.21

The Perfect Storm

Californias happy state of regulatoryaffairs changed radically in 200001 whentwo large supply shocks and one large

demand shock simultaneously hit the state.None of those shocks was triggered by statepolicy. All of them, however, had a seriousimpact on wholesale electricity prices.

The Natural Gas Price Spiral

The first supply shock was a massiveincrease in regional wholesale prices for natur-al gas, the fuel input for 49 percent ofCalifornias electricity capacity in the first ninemonths of 200022 and nearly all its peakingcapacity. In 199899 the average price of nat-

ural gas delivered to util it ies in California was$2.70 per mill ion British thermal units (Btu).23

By the summer of 2000, however, wholesalespot gas prices at the southern Californiagate had risen to $5 per million Btu, andthey had risen to an average of $25 per millionBtu by December24 (the price reached $60 permill ion Btu on December 9).2 5

The increase in natural gas prices wasnationwide. Wholesale natural gas pricesthroughout the United States duringNovember and December 2000 averaged

nearly $10 per million Btu compared to $2per million Btu during 1999.2 6 Mark Mazur,acting administrator of the EnergyInformation Administration, testified to theSenate Energy and Natural ResourcesCommittee that gas prices previously hadnot remained this high for a sustained periodof t ime.2 7

5

Two l arge supp lshocks and onelarge demandshock simul tane

ously hit thestate. None ofthose shocks watr iggered by statpolicy.


6/35

The increase in the price of natural gaswas the logical consequence of the first sig-nificant cold winter after several mild wintersand less-than-average amounts of natural gasavailable from storage.2 8

The price increase was worse in Californiabecause demand was greater relative topipeline capacity and total storage.2 9 Anexplosion in August 2000 shut down the ElPaso pipeline, which carries natural gas fromTexas to southern California. That accidentreduced pipeline capacity into the state by 10percent for several weeks, a large supply dis-ruption that has not been fully remedied.3 0

The surge in demand and the subsequentprice increase caught the industry by surprise.Natural gas prices had, after all, declined 25

percent in inflation-adjusted terms between1985 and 1999.3 1 Consumption from 1995through 1999 was essentially flat.32 Thus therewere very limited incentives to increase pro-duction or hold inventory going into 2000.

With the increase in prices, however, theoil and gas industry increased domesticexploration and development budgets by 40percent in 2000; another 20 percent increaseis expected in 2001.33 Unfortunately, newinvestments in production wil l increase out-put only after several years. The Energy

Information Administration, a semi-inde-pendent analytic arm of the U.S. Departmentof Energy, believes that prices will resumetheir historical trend by 2004.3 4The NationalPetroleum Council concurs, forecasting thatthe average wellhead price through 2010 willbe approximately $2.74 per million Btu.3 5

Given that 90 percent of a naturalgasfired generators cost of producing elec-tricity stems from fuel costs,3 6increased nat-ural gas costs must increase electricity prices.A natural gas price of $20 per mil lion Btu, for

example, translates into a production cost ofat least 20 cents per kWh for an average nat-ural gasfired plant and 32 cents per kWh forthe least-efficient power plants.37

Why Natural GasFired Electr icity

Determines Wholesale PricesThe least-efficient plants costs are rele-

vant because in commodity markets, likeelectricity, the costs of the most costly pro-ducer whose output is necessary to meetaggregate demand set the price for all theelectricity sold, even power produced fromother fuels. Thus, the California wholesalemarket cannot be understood without a fullunderstanding of the increased cost ofgasfired electricity.

This is doubly important because manycommentators have argued that, if utilitieshad signed long-term contracts with cheapersources of power, the price spiral would havebeen less dramatic. Underlying that argumentis the belief that the price of electricity in a freemarket would be a weighted average of long-term and spot prices.38 While this belief has

superficial plausibility, pricing output as aweighted average of differing prices of inputswould result in shortages.

Imagine that supermarket lettuce priceswere different depending on individual farm-ers costs. Once people realized that differentprices existed for lettuce, shoppers wouldsnap up the low-cost lettuce first. The super-market would then ask the low-cost lettuceproducer for more output at lower pricesBut could the lettuce producer easily expandoutput at the same low costs? He could if he

didnt have to pay market prices for addi-tional inputs, but that would be the case onlyif he had additional inputs under long-termcontract. And if he or other producers hadsuch spare capacity lying around under con-tract, market prices would already havedecreased to reflect the competition amongthe owners of excess supply to use the unitsfor which they had contracted.

If the low-cost producer did not havespare capacity under long-term contract andhad to pay market prices for additional

inputs like land and fert il izer, he would haveto charge higher prices for the additionaloutput to cover costs. In addition, the low-cost lettuce producer would realize thatrather than increase output at higher priceshe could raise prices on his existing low-costoutput, eliminate the shortages, make moremoney than if he used weighted-average pric-

6

Pr icing output asa weighted aver -age of d i f feri ngpri ces of i nput swould resul t i n

shortages.


7/35

ing, and not have to increase production.Electricity markets are analogous to the

lettuce example.39 Even if California utilitieshad contracted for 99 percent of their supplyat low prices, demand would exceed supply atthose prices. And if the util it ies attempted toexpand output, they would have to pay mar-ket prices for the fuel input and would losemoney on every additional sale unless theadditional output was sold at market prices.

Thus, uniform prices for all sources ofelectricity regardless of their respective inputcosts is a (good) characteristic of free marketsand not the result of the prohibition of long-term contracts in the Cali fornia system or ofthe mandate that transactions take placethrough a centralized spot market. In fact,

this very point was once made quite energet-ically by the political left to justify energyconservation subsidies. Amory Lovins, afamous advocate of energy conservation, hasargued correctly that traditional state regula-tion of the electric utility business deliveredaverage costs, not marginal costs, to ratepay-ers, which resulted in economically ineffi-cient (excessive) levels of consumption andlosses for the utilities whenever marginalcosts exceeded average costs.4 0Lovinss reme-dy for this market distortion, however, was

not a fi rst-best solutionthe introductionof marginal cost pricingbut a second-bestsolutionto have electric utilities subsidizeratepayer purchases of energy efficient appli-ances and technologies, which would becheaper for utilities than investments inadditional generation (a matter well returnto later in this paper).

The WeatherThe second supply shock was caused by a

three-year dry spell that reduced reservoir

and river-flow levels and thus reduced hydro-electric generation in California by 20 per-cent from 1998 to 1999.4 1In addition, before1998 hydropower was more abundant thannormal, reducing the returns that wouldcome from investment in natural gasfiredproduction, just as California was switchingto deregulated generation. Hydro output

in the Canadian and U.S. West in 1997 wasmore than 30 percent greater than in 1992.42

Hydropower from the Pacific Northwest fur-ther declined from an hourly average of20,805 megawatts (MW) in 1999 to 18,075MW in 2000. California hydropower likewisedeclined from an hourly average of 4,395MW in 1999 to 2,616 MW in 2000.4 3 FromJune through September 2000, hydro pro-duction throughout the West was, on aver-age, 6,000 MW less than during the samemonths in 1999, equivalent to the output of7 to 10 nuclear plants.44

Unfortunately, the water shortage willalmost certainly get worse before it gets bet-ter. Stream flows in the Pacific Northwestlast January were only about 60 percent of

average4 5

and snowpack in the CascadeRange was likewise only 60 percent of nor-mal,4 6 prompting the Northwest PowerPlanning Council to warn that hydroelectricpower generation this summer will be 5,000MW below normal.4 7 As of March 30, 2001,Californias Department of Water Resourcesreported that snow and rain accumulationwas 34 percent less than normal in thatstate.4 8 The practical effect of this reductionin hydroelectric generation was to leaveCalifornia with little spare generating capaci-

ty during peak-demand periods.In addition to the negative supply shocks

(natural gas price increase and hydro short-age), a demand shock hit during the summerof 2000 because of unseasonably warm tem-peratures (a 13 percent increase in coolingdegree-days across the Pacific region from1999 to 2000).49 Temperatures in the Arizonasubregion of the western grid averaged threeto five degrees higher than normal.5 0 TheCalifornia-Nevada subregion of the westerngrid experienced the 99th hottest summer of

the last 106 in June 2000, compared with 59thand 14th in 1999 and 1998, respectively.5 1

Accordingly, energy consumption andaverage daily loads during the summer of2000 grew rapidly compared with the sameperiod in 1999. Electricity consumption inthe western states, excluding California,increased by 4.7 percent in from May 1999 to

7

The second sup -pl y shock wascaused by a th reyear dr y spel l thareduced r eservoiand ri ver-fl owlevels and thusreduced hydro-electr ic genera-t ion inCalifornia.


8/35

May 2000, and energy consumption inCalifornia increased by 5.8 percent over thesame period. The increase in energy con-sumption from June 1999 to June 2000 waseven greater7.3 percent for the WesternSystems Coordinating Council states,excluding California, and 13.7 percent forCalifornia. Within the ISO, average dailypeak loads grew by 11 percent in May and 13percent in June compared with the samemonths of 1999.52

Even though the increase in electricitydemand may not sound dramatic, the effect onnatural gas consumption was dramatic becauseof the hydropower reductions. DuringMaySeptember 2000, natural gas consump-tion in California by utilities was 22.4 percent

greater than for the same months in 1999.5 3

Inthe West as a whole natural gas use for electric-ity generation increased an astonishing 62 per-cent during the same period.5 4

The hot summer of 2000 was then fol-lowed by a historically cold winter in 2001,ensuring that demand would remain highthroughout the region.55

Some commentators have questionedwhether this seemingly moderate increase indemand could really explain price increasesthat are severalfold more severe on a percent-

age basis than the increase in demand.56

Given the relative inelasticity of demand forelectricity in general (made worse by the exis-tence of retail rate caps, discussed below),even moderate increases in demand willincrease prices dramatically if supply is fixedin the short term, which is exactly what hashappened in the western power market.5 7

We draw two conclusions from our analy-sis of supply and demand changes inCalifornia. First, the reductions in supplyand increases in demand that resulted in

wholesale electricity price increases are theresult of natural weather variation interact-ing with market forces.5 8No state politician,regulator, or businessman could have headedthem off. Second, no regulatory systemnotthe pre-1996 regulatory regime, not the post-1996 regulatory regime, not a completelylaissez faire regime, and certainly not any of

the various regulatory regimes put in place inother statescould have prevented wholesaleelectricity prices from climbing to record lev-els under these circumstances.

Political Cloud Seeding

While we believe that the hydro shortage,natural gas price increases, weather shocksand pipeline disruptions are the proximatecauses of increased prices for wholesale elec-tricity in Cali forniathe perfect storm, ifyou willthe price increases were exacerbatedby the existence of two politically createdphenomena: nitrogen oxide (NOx) emissionpermits and retail price controls. That polit-

ical cloud seeding made the perfect stormeven more intense and unpleasant.

NOx Emission PermitsAlthough environmental regulations, in

general, affect both generating costs and theability to site new capacity, Californiasrequirement that generators have sufficientN Ox emissions credits before going onlineplayed a particularly important role in theprice spiral of 200001. To the extent thereis a smoking gun, its NOx, J. Stuart Ryan

vice president of AES Pacific, an independentpower generator, told the Federal EnergyRegulatory Commission. The cost of creditsfor NOx emissions in the L.A. Basin has sky-rocketed.5 9

It is well understood that NOx emissionsare often (but not always) an important con-tributor to urban smog in the summert ime.60

Accordingly, California adopted regulationsin 1994 (known as the RECLAIM program)that allotted a certain number of NOx emis-sions credits to existing emissions sources in

southern California and allowed them totrade those credits on the open market sothat the overall costs of emission reductionacross plants would be minimized. Plantsthat found it cheaper to buy emissions per-mits than to reduce facility emissions wouldpurchase permits from firms that found itcheaper to reduce emissions than to hold

8

No r egulator y sys-tem could have

pr event ed whole-sale electricity

pr i ces f romcli mbing to

record l evelsunder these

circumstances.


9/35

emissions permits. Each year the pool ofcredits is reduced by 8 percent. Power genera-tors must purchase enough credits to offsetemissions before they can go online or paylarge fines to the state.6 1

In the winter of 1999, NOx

credits wereselling for about $2 per pound. By the sum-mer of 2000, those same credits were sellingfor $30 to $40 per pound where they havestayed ever since.6 2 Because an efficientgasfired plant emits about a pound of NOxper MWh and an inefficient plant emitsabout two, NOx credit prices of $30 to $40per pound necessitate an additional cost of$40 to $80 per MWh ($0.04 to $0.08 perkWh).63 In January 2001, California regula-tors initiated a waiver of NOxpermit require-

ments for power generators for the next threeyears, but the damage had been done.6 4

Although its true that the RECLAIM pro-gram affects only those generators in the L.A.Basinthe source of only a fraction of thestates powerremember that the highest-costsource of electricity sets the price for allelec-tricity sold through the western grid. AaronThomas, a manager at AES Pacific, points outthat generators in the L.A. Basin are settingthe clearing price for everybody in California.And to the extent that that market is influenc-

ing markets in the West, all of a sudden youregetting these basin units driving costs for 50million people in the West.6 5

The Damage from Pri ce Control sThe wholesale electricity price increases in

California were exacerbated by the existenceof retail price controls.6 6Normally, firms thatincrease prices experience fewer sales as aconsequence. With retail price controls, how-ever, power generators could increase priceswithout fear that those price increases would

reduce revenue.Academic economists have tested this

proposition. Vernon Smith and his col-leagues at the University of Arizona have con-ducted experiments to compare the behaviorof auction prices under two scenarios: one inwhich consumers face rigid retail prices and asecond in which 16 percent of customers face

real prices that reflect supplier bids. Prices inthe second scenario are as much as 30 per-cent less than prices in the first scenario.Prices are dramatically lower as long as somecustomers face the real costs of peak-demandelectricity.6 7 Industry consultant Eric Hirstargues that if only 20 percent of the totalretail demand faced hourly prices, and as aresponse to those prices reduced demand by20 percent, the resulting 4 percent drop inaggregate demand could cut hourly prices byalmost 50 percent.6 8

In addition to the lack of demand respon-siveness created by the price controls, the ISOplaces an infinite value on keeping the gridup, which exacerbates the problem. Aswholesale prices increased after May 2000,

the ISO enacted price controls in the marketfor daily backup and load-following (ancil-lary) reserves. On June 28, 2000, prices werelimited to 50 cents per kWh and on August 7,2000, to 25 cents per kWh.69 On November 1FERC issued a soft price cap of 15 cents perkWh for both the California day-aheadpower exchange and the real-time and ancil-lary ISO markets.7 0Given our analysis of theeffects of the prices of natural gas and NOxpermits on the marginal costs of the least-efficient natural gasfired generators, the

price ceilings had the practical effect of dri-ving backup and load-following peak-demand generators out of the market.

FERC data support this conclusion. Thecommission reports that out-of-state genera-tors reduced their sales to California fromMay through August 2000 relative to Maythrough August 1999 by an average of 3,000MW a day and their sales to the daily marketoperated by the ISO by 2,000 MW a day.71

Tom Williams, an executive at Duke Energy,confirmed that those price caps discouraged

the company from selling power inCalifornia during peak-demand periods, fur-ther shortening supply on the day-beforemarket.72

But the ISO still had the responsibility tokeep the grid operating and had to ask thosesame generators to supply on an emergencybasis regardless of price. In January 2001, for

9

I f onl y 20 percenof the total retaidemand facedhour ly pri ces, thresul t ing 4 per-cent drop inaggregatedemand couldcut h ourl y pri ces

by almost50 percent .


10/35

example, the ISO paid $461 million for last-minute purchases whereas in January 2000the ISO spent only $2.7 million on last-minute grid support. The price controls com-bined with unlimited willingness to pay foremergency power on the part of the ISOinduced high-cost suppliers to stay out of theday-before market operated by the powerexchange and hope for emergency calls fromthe ISO at very high prices. A vignette in theNew York Timesil lustrates the problem:

With its employees focused solely onkeeping the lights onand with thebills to be paid not by itself but bythe ut ilitiescritics say the agencywas ripe to be outfoxed by power

suppliers.My people are not trained astraders or arbitrageurs, concededJim McIntosh, director of schedulingfor the system operator.

When traders were asking $2,000for a megawatt-hour of electricity,enough to light 1,000 typical homesfor an hour, Mr. McIntosh and oneof his bosses refused.

But the price they settled on$1,500was arbitrary, Mr. McIntosh

acknowledged.7 3

Moreover, as the fiscal plight of the utili-ties became public knowledge (a plightentirely caused by retail price controls thatprevented the utilities from raising enoughrevenue to pay for electricity from the statepower exchange), some generators left themarket out of fear that the IOUs they werereceiving from the utilities would never berepaid. The California ISO, for instance,reports that the March blackout was largely

caused by generators shutting down 2,000MW of production because they had notbeen paid for the power they sold to the util-ities for three or four months.7 4

Other generators reflected the risk of non-payment in the prices they charged the state.Gary Ackerman, executive director of theWestern Power Trading Forum (a trade

group representing western power genera-tors), forthrightly acknowledges that thepolitical and financial uncertainties of doingbusiness in California have driven up theprices that forum members charge forpower.7 5 Duke Energy confirms Ackermansassertion, acknowledging that the prices itcharged Cali fornia uti li ties included a creditpremium to cover the risk of nonpayment.76

Pacific Gas & Electric agrees. The compa-ny alleges that on January 18, for instance, sixsuppliers that had provided 36 percent of itsdaily supply had stopped service to the utili-ty or threatened to do so. PG&E later lostmore suppliers that had provided another 30percent of supply.77 And on March 19 abouthalf of the capacity of independent power

producers (3,100 MW) who had contracts tosupply the incumbent utilities was shutdown because of nonpayment.78 In light ofPG&Es bankruptcy on April 6, 2001, thatbehavior seems very prudent.

Explaining the BlackoutsIn a free market, reductions in supply do

not result in shortages because pricesincrease enough to equate supply anddemand. Higher prices induce firms to createadditional supply as quickly as possible to

obtain profits. Eventually, the high priceseliminate the shock.

Wholesale electricity prices in Californiasince the spring of 2000 have averaged morethan 15 cents per kWh.7 9 Retail electricityprices, however, were capped at less than 7cents per kWh. Wholesale prices signaledthat electricity was increasingly scarce, butretail prices told consumers that nothing hadchanged. Accordingly, consumers demandedmore electricity than was available. Blackoutsare the inevitable result of attempts to regu-

late and legislate basic economic laws out ofexistence.

Unexpected shortfalls in transmission capaci-ty also have contributed to the blackouts. Themain transmission line between northern andsouthern California (Path 15) has at times beenlimited to half its normal capacity, significantlycomplicating northern Californias power supply

10

The pol i ti cal andf inancial uncer -

taint ies of doi ngbusiness in

Cal i forn ia havedri ven up t he

pr ices that gener-ator s charge for

power.


11/35

thus the rolling blackouts have been confinedlargely to that region.8 0

Market Manipulation and

Price GougingDo supply and demand shocks fully

explain high California electricity prices?Industry analyst Edward Krapels believes thatsupply and demand shocks explain theDecember 2000 wholesale price in Californiabut do not explain 5 cents per kWh of the aver-age April through November price.81 MITsPaul Joskow and economic consultantEdward Kahn also conclude that high whole-sale prices observed in summer 2000 cannot

be explained as the natural outcome of mar-ket fundamentals in competitive marketssince there is a very significant gap betweenactual market prices and competitive bench-mark prices . . . high prices experienced in thesummer of 2000 reflect the withholding ofsupplies from the market by suppliers. 82

Where Did Al l the Power Go?

Californians served by the ISO typicallydemand 45,000 MW of electricity duringpeak periods in the summer, and that elec-

tricity was available during the summer of2000. By contrast, Californians served by theISO demand only about 30,000 MW duringpeak periods in the winter. How could 15,000MW of power disappear and result in black-outs during the winter of 200001?

One answer, of course, is that it was soldto other markets in which wholesale pricecaps are not in place. The $250 per MWhwholesale price cap established by the ISO inAugust was below the marginal costs of theleast-efficient natural gas units with the

greatest NOx emissions in the Los AngelesBasin. A number of power generators simplystopped producing as soon as the cap wasput in place.83

Another possible explanation is that thehydropower that was lost during the summerof 2000 was replaced by many old natural gasunits that postponed needed maintenance

and repairs until this winter.8 4Approximate-ly 65 percent of the states generating capaci-ty, after all, is in plants 35 years old or older.85

This is particularly the case with the plantsused to meet summer peak demand. Thosegasfired plants seldom operated steadilyuntil the summer of 2000, and, because theyare older and more inefficient, they requireadditional repair and upkeep. At varioustimes during the winter of 200001 powerplants were out of service in unprecedentednumbers.8 6 During the first blackout onJanuary 17, for instance, fully 11,000 MW ofin-state capacity was offline for repair andmaintenance work.87

Disbelief of the repair-and-maintenanceargument is rampant. Harvey Rosenfield and

Doug Heller of the Foundation for Taxpayerand Consumer Rights call for state agents toobtain search warrants and subpoenas toenter the power plants to determine the truecause of the shortages. If necessary, theplants should be seized to protect the publichealth and safety.8 8 Of course, its unlikelythat such investigations would unearth theequivalent of a smoking gun. After all, com-panies are more than capable of justifyingmaintenance work. An investigation byFERC found that generating outages in

California at plants owned by Dynergy, NRG,and Reliant appeared to stem from increaseddemand and age of the units (boiler tube andseal leaks, turbine blade wear, valve andpump motor failures, etc.).8 9

Economists are divided about the role ofmarket power in explaining the behavior ofCalifornia electricity markets.9 0 Using pub-licly available estimates of the heat rates ofelectric generators and natural gas price data,Severin Borenstein, James Bushnell, andFrank Wolak from the Universities of

California-Berkeley and Stanford generatedestimates of the marginal costs of electricity.They then compared those estimates withactual prices observed on the CaliforniaPower Exchange. They argue that the exerciseof market power raised California electricityprices at least 16 percent above the competi-tive level (marginal cost) from June 1998

11

Supp ly anddemand shocksexplain theDecember 2000

wholesale price iCali f orni a but dnot explain 5cent s per kWh ofthe average Apr ithroughNovember pr ice.


12/35

through September 1999.9 1 Joskow andKahn studied the summer 2000 Californiamarket and concluded that an inordinatenumber of plants were taken offline formaintenance when the price spikes weremost intense. While that could be coinci-dence, it would also be perfectly consistentwith market power: the withholding of someunits of production to raise the price of elec-tricity produced by other units owned by thesame firm.9 2

An Alternative Explanation: RegulatoryPerversity

But Harvards Bill Hogan and ScottHarvey of LECG, a consulting firm, arguethat generators in California priced above

marginal cost, not because they had marketpower, but because two characteristics of theCalifornia market created incentives forthem to place high bids.9 3 The Californiamarket solicited hourly bids from all genera-tors a day in advance of production. In addi-tion, once the day-ahead market cleared andproduced hourly prices, the ISO asked forbids for reserve power. Under ideal condi-tions, arbitrage would result in similar pricesfor day-ahead and reserve energy in eachhour, but because the ISO auction was con-

ducted after the power-exchange auction,uncertainty existed about whether a genera-tor would receive a higher price if it waitedfor the ISO to call on a unit as a reserve unit.This sequential feature converted the day-ahead auction from an everyone-gets-the-market-price auction (and thus generatorswould bid at marginal cost to ensure beingselected to produce as long as costs were cov-ered) to a pay-as-you-bid auction in which allparticipants, if selected, receive what they bidrather than the (potentially) higher market-

clearing price.Under such a market structure, firms

lacking market power will bid at expectedmarket-clearing prices rather than at margin-al cost.94 And market-clearing prices for thereserve market during a capacity shortage inwhich retail consumers do not face highprices will be very high indeed because under

the engineering procedures developed by theNERC, electric system operators such as theISO must maintain generation reserves avail-able in 10 minutes regardless of cost.95

Producers have no reason not to name astratospheric price for their power in thosecircumstances because the price they chargewill not alter demand (remember, retailprices are capped and, even were they not,ratepayers face lagged-average rather thanreal-time marginal prices for electricity). AsUniversity of California-Berkeley economistSteven Stoft puts it, The failure of con-sumers to respond [because of price controls]is the fundamental flaw that makes pricesreach exorbitant levels when there is a littlescarcity or when suppliers have even a little

market power.9 6

University of Maryland and Resources forthe Future economist Tim Brennan arguesthat another characteristic of the Californiaauction market also created incentives forhigh rather than low market-clearing prices.97

The rules of the auction allowed generators tooffer different amounts of electricity at differ-ent prices rather than all of their output at oneprice. Under those rules, generators had theincentive to offer a small amount of their out-put at very high prices because, if the high bid

was accepted, they would receive that price forall their output. And if the bid was not accept-ed, the generators would lose only the sale of asmall fraction of their possible output.Normally such bidding behavior would beunprofitable because the probability of thehigh bids being accepted would be small, butin a very tight supply situation, the probabili-ty of the bids being accepted rises consider-ably, and the opportunity cost of the unsoldpower falls.

Price Gouging in Perspecti veEven though perverse behavior by electric

generatorsinduced either by actual with-holding or by characteristics of the Cali forniaauctionmay have played some role in theprice spike, the populist charge that theentire price spike can be explained by pro-ducer manipulation is clearly nonsensical.98

12

Generators hadthe incent ive to

of f er a smallamount of t heir

output at veryhi gh pri ces

because, i f thehi gh bi d was

accept ed, t hey

woul d r eceive thatpri ce for al l t hei r

output.


13/35

Input costs and natural scarcities are respon-sible for most of the price hike. Some of thetop economists in the electricity establish-mentindividuals on both sides of the dis-puteagree that, although market manipula-tion may be present, it is not the fundamen-tal cause of the crisis.9 9

Consider the data available on wholesaleprices in January and February 2001. FERCmaintains that it cost 27 cents, on average, toproduce a kilowatt-hour of electricity in nat-ural gasfired generators during stage 3emergencies in California last January.100 TheCalifornia ISO reported, however, that it paidan average of just over 28 cents for powerpurchased in the real-time spot market.101

During February, wholesale natural gas

prices increased dramatically, resulting inFERC estimates of production costs of 43cents per kilowatt-hour during stage 3 emer-gencies.102 Unfortunately, there are no pub-lished data that allow a comparison of theFERC estimate and actual prices observed inthe real-time spot market, but if the Januaryrelationship held in February, wholesaleprices were not more than a few cents higher,on average, than production costs.

To be sure, industry analysts believe thatactual production costs during power emer-

gencies are far higher than FERC believeslargely because FERC relies on a monthlyaverage for wholesale natural gas prices thatinadequately reflects day-to-day pricechanges,103 but at least FERC has recognizedthat input prices and the NOx quantityrestrictions explain most of the price increaseand that the highest-cost source of supplynecessary to meet demand sets the legit imateprice for all power sold in the market.104

Big Profits? Big Deal

The evidence most commonly used to jus-tify charges of price gouging is the huge prof-its earned by many independent power gen-erators doing business in California.California Senate president pro tem JohnBurton, for instance, supports his charge ofprice gouging by noting that power genera-tors and power marketers have seen their

profits rise by an average of 508 percent overthe past year.105But large profit margins arenot themselves evidence of prices above mar-ginal costs of the most-expensive unit.

The key to understanding the electricitymarket is, again, to remember that the high-est-cost source of supply necessary to meetdemand sets the price for all electricity in themarketplace. Thus, inframarginal genera-torssuch as those who have access to cheapgas through long-term contracts or who havehighly efficient power plantscan and do sellat the market-clearing price and make profits(a state of affairs that is found satisfactory byboth FERC and the California ISO).Marginal generatorsthose who have noaccess to cheap gas or who operate old, inef-

ficient plantshave no such market advan-tages and thus fewer profit opportunities.The profits reported in the media largelyreflect the profits made by inframarginal, notmarginal, producers.

Are Environmentaliststhe Culpr i t?

Many observers have argued that theCalifornia electricity shortage is the result of

environmentalists and consumer activistsefforts to block new generation and transmis-sion capacity, slowly starving the state of need-ed power.106 Californias reserve generatingcapacity decreased from 40 percent in 1990 to15 percent at the beginning of 2000.107 Lastsummer, demand for electricity was up 23 per-cent compared to 1992, yet generating capaci-ty had grown by only 6 percent.108

A Murder with No Body?Although its certainly true that California

has not seen a boom in power plant construc-tion over the past decade, the claim that nonew power plant has been built in Californiain more than 10 years is utterly false.According to the California Energy Commis-sion, 11 power plants (10 gas fired, 1 coalfired) with a generating capacity of 1,206 MWbegan operation in California in the 1990s.109

13

Al though mark emanipulationmay be pr esent , i s not th e f unda-ment al cause ofthe cr isis.


14/35

Only two licensed projects with a generatingcapacity of 229 MW were dropped in the1990s (one because of bankruptcy), and onlyone project was blocked by communityactivists: a 240 MW set of barge-mounted gen-erators that were to supply peaking capacity tothe San Francisco Bay area.

Opposition to new power plants is notconfined to environmentalists, and someenvironmental groups actually support newplants. The Calpine Co. has proposed build-ing a 600 MW gasfired plant in San Jose.Many environmental groups, including theSierra Club, actively support the plant. Themain opponent, ironically, is the Internetgiant Cisco Systems, which fears that a powerplant near its facilities would be aesthetically

unpleasing and thus reduce its ability toattract employees.110

Not only were environmentalists a relativenonfactor in generator investment decisionsin the early to mid-1990s, they scarcelyplayed any role in blocking new capacity inthe months leading up to the crisis. SinceGovernor Davis was elected in 1998,California has approved the construction of9 power plants, and 44 plants with 22,600MW of generating capacity are currentlyunder consideration by the California Energy

Commission.111

All of them will almost cer-tainly be approved, but they will take severalyears to build.112 Tens of thousands of addi-tional megawatts of capacity are under con-struction in other states that are part of thewestern regional electricity grid. A new studyby Resource Data International, a division ofFinancial Times Energy, concludes, Even inthe West, where shortages and unprecedent-ed high prices have been the rule in 2000,more than enough new capacity is underdevelopment to bring power markets into

balance and perhaps provide a mild over-cor-rection within the next couple of years.113

One could argue that investors fear ofenvironmental opposition and the costs offighting that opposition to a successful com-pletion explain the lack of plant constructionin the 1990s, but little evidence exists to sup-port that claim.

More obvious explanations for the lack ofinvestment in new capacity are low prices andregulatory uncertainty. As discussed earlierwholesale power prices in California were solow that there was little profit to be gained byincreasing production. William Keese, chair-man of the California Energy Commission,explains that the demand for generation thatbuilt up through the 1990s was for needlepeak demand, defined as the additional sup-ply that would be needed to meet a 4,000 MWsurge on the third consecutive day of record-high temperatures. Such peaks are rare (31 inthe last 40 years), making new peaking capaci-ty difficult to pay for. To say people shouldhave built power plants is not rational becausethey would have lost money.114

Moreover, regulatory uncertainty keptinvestors out through much of the mid-1990s. No power plant applications werefi led with the Cali fornia Energy Commissionfrom 1994 to 1998 because of the investoruncertainty created by deregulation.115

Low prices and regulatory uncertaintyalso explain why states in which environmen-tal activism is low experienced no moreinvestment in new power plants thanCalifornia. Arizonas population grew 40 per-cent in the 1990s. The state has a pro-busi-

ness climate, and yet its power productionrose only 4 percent during the 1990s, mostlyat existing plants. No one applied to build amajor plant in Arizona between the late1980s and late 1999.116

If California had built more power genera-tors during the 1990s, they would almost cer-tainly have been gasfired facilities becausethose were the cheapest and easiest plants tosite. And because the electricity price increaseis largely a reflection of the regional increase inthe price of natural gas, those hypothetical

plants would have reduced California pricesonly if enough plants had been built so thatthe new plants (with their lower heat rates)became the marginal source of electricity(rather than older natural gas plants withhigher heat rates) and enough plants had beenbuilt to eliminate any scarcity rents that arecurrently in California prices.

14

Obvious explana-ti ons f or t he lackof investment innew capacit y are

low pr ices andregulatory

uncertainty.


15/35

In short, given the increase in natural gasprices observed in California, a massiveinvestment in natural gas plants would stillhave resulted in 15 cent per kWh electricity.And if retail consumers were still paying 6.7cents per kWh, bankruptcies and blackoutswould stil l occur.

Environmentalist and NIMBY (Not in MyBack Yard) agitation, however, has sloweddown the regulatory approval process, whichrequires much more time in Cali fornia than inother states (several years versus severalmonths). A running joke reportedly popularat the California Energy Commission is thatno plant gets approved unless the paperworkweighs as much as the plant itself. 117 Theinability to respond quickly to the supply

shock with new power generation can fairly beblamed on rules written to please Californiasactivist community, but again, the Greens arebut one part of that community.

Did Ki ll ing the Nukes Ki ll Cali fornia?

Californias well-known campaign againstnuclear power has also been blamed for thecrisis. In 1989 the nuclear reactor at RanchoSeco was prematurely decommissionedbecause of environmentalist agitation, sacri-ficing 913 MW of power daily. In 1992 the

San Onofre Unit-1 reactor (one of three reac-tors on the site), generating 436 MW ofpower daily, met the same fate.

Blaming the crisis on Californias anti-nuclear activists, however, makes little sense.First, even states without well-organized anti-nuclear organizations and growing electricitydemand failed to attract investment innuclear power, suggesting that other factorswere at work in discouraging investment innuclear power plants. Second, naturalgasfired plants throughout the 1990s were

simply better investments than nuclearpower plants; they were quicker and cheaperto build and, with fuel costs so low, not thatmuch more expensive to operate.118 Evenwithout state prohibitions and environmen-talist opposition, its extremely unlikely thatinvestors would have chosen to build nuclearplants instead of gasfired generators.

Similarly, if the Rancho Seco and the SanOnofre Unit-1 reactors had been kept open,prices would be lower today only if the extranuclear capacity reduced the number ofhours in the day for which natural gas unitsset the market price. But neither hydro nornuclear is the costliest source of power dur-ing the daytime, and thus they have no effecton daytime prices. Additional nuclear capac-ity could have lowered off-peak prices if ithad been used instead of natural gas unitsduring off-peak hours.

We repeat: wholesale prices are driven bythe costliest source of power at the margin,so the price shock would have been morebenign had those nuclear plants been onlineonly if they had displaced the use of natural

gas during off-peak hours. An additionaldaily supply of 1,400 MW might well haveheaded off some of the blackouts, butbecause transmission constraints into north-ern California play such a large role in thoseblackouts, only the 913 MW from RanchoSeco, located near Sacramento, would havehelped.119

How Green I s the Grid?Concern only about Californiasgenerating

capacity, however, ignores the fact that the

electric power market in the West is one large,interconnected system. There is no reason todemand that California internally generate allits power any more than to demand thatRhode Island produce all the food it con-sumes. Generation not built in San Franciscofor Bay Area ratepayers, for example, simplygets built elsewhere in Utah, Arizona, orNevada and is sent to the Bay Area over wire.

But the transportation of electricityrequires transmission capacity. Even thoughtransmission capacity costs only one-tenth as

much as generation, landowners and otherlocal residents resist new transmission capac-ity.120 From 1989 through 1997, transmis-sion capacity per MW of summer peakdemand declined by 16 percent. Between1997 and 2007, transmission capacity rela-tive to summer peak demand is expected todecline another 13 percent.121

15

The inabil it y torespond quicklyto t he suppl yshock wit h newpower generat iocan fai r ly beblamed on r ul eswri tt en t o pleaseCalifornias

activistcommunity.


16/35

But the problem of transmission con-straints exists all over the country, not just inCalifornia. While environmentalists havebeen known to agitate against new powerlines, theyre scarcely the onlyor even thelargestgroup of NIMBY-ites to do so.

Moreover, the lack of new transmissioncapacitylike the lack of new generation capac-ityhas as much to do with economics as withpolitics. Incumbent utilities have little incentiveto build new capacity that would make it easierfor ratepayers to buy cheaper power from com-petitors in neighboring states.122 And utilitiesdo not have an incentive to invest in new capac-ity when the profits allowed them by regulatorsare too low to make those investments particu-larly worthwhile relative to unregulated invest-

ments.123

And with transmission rules still upin the air and unsettled at both the federal andstate level, regulatory uncertainty is also damp-ing investment.124

News stories also claim that natural gaspipeline capacity constraints are the productof environmentalist or NIMBY-ite opposi-tion.125 But there is little evidence to suggestthat investors have been inhibited fromincreasing pipeline capacity when profitopportunities presented themselves. Extensivenew pipeline capacity into northern California

from Canada, for example, was built in the1990s.126 The Energy Information Adminis-tration observes that pipeline capacity hasgrown with end-use demand, and as new sup-plies have developed, new pipelines have beenbuilt to bring this gas to markets.127The GasResearch Institute likewise finds that growthin pipeline capacity is not a constraint ongrowth in gas supply. . . . If supply is available,history has demonstrated that the pipelineswill be built as needed. It is simply an invest-ment and engineering issue.128

Little pipeline capacity into southernCalifornia was added during the past decadebecause investors found few opportunitiesfor profit in the construction of newpipelines. The existing pipelines were notfully utilized until this year.129High naturalgas prices, however, have revived interest inpipeline capacity expansions, and three sig-

nificant projects were recently announced totake advantage of the newly discovered prof-it opportunities in transmission.130 Clearlythe barriers to pipeline expansion in Califor-nia are not too terribly high when profitopportunities exist.

Accordingly, its hard to single out theenvironmentalists as the cause of transmis-sion constraints. While theyve certainlyplayed a role in opposing grid expansion,even states without well-organized environ-mentalist lobbies have found it difficult toremedy transmission congestion.

Did Other States Adopt Better Designs

Than Cali fornia?Many analysts have argued that other

states deregulated more intelligently thanCalifornia and thus did not experience thesame high prices and shortages.131After allthe average price for wholesale power inCalifornia was $313 per MWh in the middleof January 2001 compared to $74 per MWhin New England, $63 per MWh in New York,and $39 per MWh in Pennsylvania, NewJersey, and Maryland.132

Fuel I nputs Explain Al lAlthough other states have restructured

their electricity regulations differently thanCalifornia has and have not experienced largewholesale price increases, the characteristicsof the state plans are not responsible for thelower prices. Pennsylvania, for example, didnot force utilities to sell off nearly as manygenerating assets as did California andallowed long-term contracting for power andestablished a more robust retail and whole-

sale market of competing suppliers.133 Butthe price differential between power inCalifornia and power in the other states thatrestructured their electricity regulations isthe result of fuel composition, not of a bet-ter regulatory climate.

States east of the Rocky Mountains do notrely on natural gas for their electricity during

16

The lack of newtransmission

capaci tyl ike thelack of new gener -

at ion capaci tyhas as much to dowi th economics as

with pol it ics.


17/35

the winter (and only a few rely heavily on nat-ural gas during peak-demand periods in thesummer). For example, during the first ninemonths of 2000, 19 percent of electricity wasgenerated by natural gas units in NewEngland, 18 percent in New York, 49 percentin California, and only 4 percent in Pennsyl-vania, New Jersey, and Maryland.134 Easternstates reliance on nuclear, coal, and hydro-power (which in the East has not sufferedfrom drought) in the winter explains why theelectricity crisis has been confined thus faralmost exclusively to California and neighbor-ing states, which have had to substitute natur-al gas for hydro in the production of electrici-ty in the winter.

If major exogenous supply and demand

shocks were to hit the states touted as dereg-ulatory successes, they, too, would findthemselves experiencing an increase inwholesale electricity prices. Blackouts wouldprobably not occur, however, because, as bestas we can tell, all states except California havefuel pass-through provisions that allow ris-ing fuel costs to be incorporated sooner orlater into electricity rates.135

Is Coal Better Than Gas?In light of the above, some people might be

tempted to argue that California went wrongby not embracing coal generation. A typical coalplant is more expensive (almost four times) tobuild per kW of capacity and less efficient tooperate (more Btu from coal are required togenerate a kWh of electricity than Btu from nat-ural gas in a new combined-cycle plant), but thecost of coal per million Btu of heat output isless than that of natural gas.136 Thus compar-isons of the two technologies require a compar-ison of capital and expected fuel costs at expect-ed utilization rates. In 1996 two EIA analysts

determined that under all reasonable scenariosnatural gas plants are more cost-effective thancoal.137 The EIA again echoed that finding lastyear in Annual Energy Outlook 2001:Natural gastechnologies tend to dominate projected capac-ity additions because the total cost of electricitygeneration from these technologies is less thanthe other options.138

Thus, California regulators cannot befaulted for not encouraging additional coal-fired generation. Investors, not state regula-tors, determine what sort of power plants arebuilt, and investors acted rationally byeschewing coal-fired facilities in favor of nat-ural gasfired facilities.

Reservations about Reserve CapacityCan Californias regulatory regime be

blamed for providing less reserve capacity thanis maintained in other states? Does someoneneed to think about reserve capacity, or canthe choice be left to market forces?139

New England, New York, and the mid-Atlantic states have installed-capacity require-ments in their deregulated markets. Utilities

are required to have available generationcapacity to meet their peak demand or facepenalties. New York, for example, sets thepenalty at three times the estimated averagecost of building a peak-demand unit, whichpresumably induces utilities to build or con-tract for sufficient peak-demand units.140

California and Ontario do not have installed-capacity regimes. When supplies are tight inthose regions, prices will rise sufficiently toinduce the installation of new supply.

Retail power prices in the installed-capac-

ity regimes are higher, on average, to pay forthe excess capacity, but peak prices are prob-ably lower.141 But damped peak prices, ineffect, socialize the costs of meeting peakdemand and induce less consumption reduc-tion by consumers during peak periods, bothof which are economically inefficient. Theinstalled-capacity regimes are a holdoverfrom the cost-of-service regulated era with itsbias toward new supply rather than demandreduction through the use of price signals.

In an unregulated market, excess capacity

would lead to prices equal to marginal cost.Because electricity generation is very capitalintensive (thus fixed costs are high relative tomarginal costs), firms would not recovertheir fixed costs under marginal cost pric-ing.142 Thus excess capacity would be builtonly if subsidies were offered or regulationsenacted rather than as the result of simple

17

Eastern statesreliance onnuclear , coal , anhydropower in

the win terexplains why theelectr ici ty cr isishas been con -f i ned almostexclusively toCal if orni a andneighboringstates.


18/35

market forces. Therefore, unregulated elec-tricity prices will be more volatile than pricesin other markets in which marginal costs arenot much lower than average costs. ThoughCalifornia may appear to provide evidence ofthe failure of the market system and supportfor the installed-capacity regime, that is trueonly because of retail price controls, whicheliminate demand reduction when suppliesare tight and exacerbate the price increasesobserved in the wholesale market.

The Di fference between California andthe Rest of the West

Even though electricity regulation andpolicy discussion occur at the state level, theUnited States really has three electricity sys-

tems: the eastern interconnection, the west-ern interconnection, and the Texas system,which has limited direct- rather than alter-nating-current connections with the easterninterconnection.

The focus on California creates theimpression that the other states in the west-ern interconnection are exempt fromCalifornias difficult ies. That is only partiallycorrect. Because electricity supplies arebought and sold all over the western inter-connection, in the absence of transmission

constraints, transportation-cost-adjustedprices everywhere in the West will be thesame.143 Thus it is not possible for any west-ern state to insulate itself from a shock tosupply or demand that occurs anywhere inthe western interconnection.144

Even though the transmission grid doesnot allow states to insulate themselves fromeach others electricity situations, the percep-tion exists that other western states built moresupply than did California in the 1990s rela-tive to demand increases and thus are shoul-

dering more than their share of responsibili tyfor electricity supply.145 That is incorrect.

Demand has been exploding outsideCalifornia, and supplies have not increasedcommensurately. Eighty-five percent of thegrowth in electricity demand in the West since1995 has occurred outside California.146 Butno state except Montana has increased electric

power production more than populationgrowth. In the 1990s Arizonas populationgrew by 40 percent and electricity demandgrew at nearly double that rate, but power pro-duction rose only 4 percent despite Arizonashaving almost none of the environmentalrestrictions blamed for keeping Californiafrom expanding capacity.147 In Nevada duringthe same period, population grew by 66 per-cent but power production by only 44 per-cent.148 John Harrison, a member of theNorthwest Power Planning Council, whichrepresents Idaho, Montana, Oregon, andUtah, said: Our supply is extremely tight tooWe havent been building any new powerplants either.149

The main difference between California

and other western states and the explanationfor the lack of crisis mentality in the otherstates is that the latter have increased prices toconsumers whereas until recently Californiahas not. Nevada will have raised residentialrates by 75 percent from 2000 to 2003.150

Rates in Tacoma, Washington, have risen 50percent, and utilities in Idaho, Arizona, andUtah have also announced large increases.151

These price increases are average rather thanreal-time, but at least prices are rising. Thusutilities do not go bankrupt, and suppliers do

not withhold because of fears of nonpayment

The Rescue Plans

A number of initiatives have been under-taken to alleviate the power crisis inCalifornia, and other plans remain on thedrawing board. Unfortunately, few if anyaccurately diagnose the problem or treat any-thing beyond the superficial economic symp-toms of the underlying disease.

Price ControlsGov. Gray Davis and most of the

California political establishment believethat political management of supply shocksis better for consumers than economic man-agement via free prices. Their argument issimple: When supply and demand are fairly

18

The main dif fer -ence between

Cal if orni a andother western

states is that t helat ter have

increased pr icesto consumerswhereas un t i l

recentlyCal i forn ia has

not.


19/35

rigid in the short runin this case, because ofboth natural market characteristics and bur-densome regulatory policieshigh prices willover the short run mostly transfer wealthfrom consumers to firms rather than inducenew supply or decrease demand. This is par-ticularly true for baseload coal and nuclearplants, the marginal costs of operation ofwhich are much less than those of the oldinefficient natural gas units used to meetaggregate demand in California.152 So gov-ernments institute price controls and issuelegal orders to require supply increases anddemand reductions rather than allow pricesto achieve the same result.

What if regulators attempted to reducewholesale prices by restricting the profit

opportunities of inframarginal supplierswith, say, a windfall profits tax or some sortof cost-plus rate cap? Negative supply anddemand effects would likely result.

On the supply side, such policies wouldreduce the incentive to invest in cheap energyproduction. After all, why care whether yourplant is more efficient or cost competitivethan the competition if the state ensures thatyour profit margins are no greater than thoseof your less-efficient competitors? Such poli-cies would also bias investment away from

facilities with high up-front costs and lowmarginal costs (coal, nuclear, and mostrenewables) and toward facilities with lowerup-front costs and higher marginal costs(natural gas and fuel oil). Both fuel diversityand economic efficiency would suffer.

On the demand side, mandating pricesbelow market-clearing levels would send thewrong price signals about the cost of con-suming additional units of electricity, artifi-cially increasing demand at the very t ime thatdemand reductions are necessary to prevent

collapse of the grid into blackout status.Such a demand increase would have the per-verse effect of actually increasingthe price ofpower from marginal suppliers (more con-sumers chasing a fixed amount of service).

Can polit icians intervene in the electricitymarket to reduce the wealth transfer and yetpreserve the important incentive functions of

prices? In theory, governments can enactprice controls on inframarginal (existing)output but leave incremental supply freefrom controls. This preserves static efficiencyby pricing incremental demand and supplycorrectly and reduces the wealth transferfrom consumers to producers. But, to theextent investment occurs because of the pos-sibility of making profits during periods oftight supply, dynamic efficiency is negativelyaffected by even this policy. In addition, oncethe government controls prices on existingoutput, political struggles ensue over accessto that output. As economist Ben Zycherwarns, Feel free to bet the rent money on theprospect of politicized electricity rates,designed to subsidize various consumer

groups and geographic regions with impor-tant allies in the Legislature at the expense ofother groups.153

Consider two past policy schemes that areanalogous to the schemes that GovernorDavis proposes for California: the federalintervention in oil markets after the 1973 oilshock and rent control in New York City. TheEmergency Petroleum Allocation Act of 1973allowed new supplies of crude oil andincreases from old supplies to commandmarket prices but controlled prices for exist-

ing supplies. Because of historical patterns,some refineries had 0 access and others 100percent access to old oil, which caused enor-mous variation in profits because gasolinesold at market-clearing prices, which werehigh regardless of the amount of old oilused in the refining process. That, in turn, ledto a political fight over access to the wealthcreated by price-controlled old oil. Theonly winners were the lawyers who wrote theregulations and tried the cases about whatwas and was not old oil and refineries that

had access to cheap old oil.154

Similarly, in New York City after WorldWar II rent controls were left on the existinghousing stock, but new units were unaffect-ed. The city has since reneged on its pledge tonot impose controls on units built after1947, and the controls have led to a blackmarket in which access to rent-controlled

19

Once the government contr olspr ices on exist inoutput, poli ti cal

st ruggles ensueover access to t houtput.


20/35

units is sold with the proceeds going toexisting tenants and brokers rather than tothe legal owners.155

In both the oil and housing policy experi-ments, the wealth transfers suppressed byprice controls were not recycled to the publicat large. They were instead captured by oilrefineries and lucky tenants who happened tolive in price-controlled housing. And becauseof investors concerns about future confisca-tion of profits, investment has been less thanoptimal, particularly in NYC housing.

Legal Orders

Legal orders can reduce demand, but atbest they are much more costly to developand enforce than price increases, which lead

all people, not just government lawyers, tothink about how to conserve. Even if theywork, many orders have only symboliceffects. And at worst they have perverse eco-nomic consequences.

News articles often describe the manychanges in behavior that mangers proposeafter a legal order is issued. For example, inresponse to an order by the governor to allstate departments to reduce energy use by 5percent, the California Department of WaterResources announced that it would use aque-

duct pumps at night rather than during thedaytime peak.156 The implication is thatorders are effective. But under real-time mar-ket prices, use of aqueducts off peak, as wellas many other changes no one has everthought about, will occur without orderstelling people what to do.

The category of symbolic but ineffectivelegal orders is well represented by GovernorDaviss proposed regulation to require cardealerships to turn off most of their lights atnight.157 But electricity is relatively cheap and

plentiful at night. Its electricity during thedaytimeparticularly on warm summerafternoons and early eveningsthat is expen-sive and scarce. Nighttime conservation doesnothing to prevent blackouts.

The perversity of regulation is superblyillustrated by a proposal to ban power produc-ers from exporting electricity to other states.158

Such a regulation would lead to other statesrestricting electricity exports to CaliforniaAnd because California typically importsmuch more power annually than it exportssuch a trade war would have devastating con-sequences for California ratepayers.159

Subsidizing Conservation

Desperate to reduce electricity demand bysome mechanism other than the eliminationof retail price caps, California has adopted an$850 million energy conservation initiative,which consists primarily of programs to sub-sidize ratepayer purchases of energy efficientappliances and technologies.160

The fundamental problem with such sub-sidies in a price-controlled market is that

energy efficient appliances reduce the costsof operation. That may not be a major issuewhen it comes to, say, the television set (wewont watch more TV just because it costs alittle less to turn on the set), but for appli-ances like air conditioners that make all thedifference during peak-demand periodsenergy efficiency simply reduces the margin-al cost of energy services and thus increasesnot decreasesenergy consumption. This isknown as the rebound effect, and its beenwell established by economic researchers.161

The upshot is that the gains of energy effi-cient appliances are at least partially offset byincreases in appliance operation time.

Its not as if we havent tried such pro-grams before. Uti lities across the nation havespent about $20 billion since the mid-1980sto subsidize ratepayer investments in energyefficiency.162 Yet examination of the datareveals that utilities that invested heavily insubsidies experienced no reduction in elec-tricity demand relative to utilities that didnot invest in subsidies.163

There are several reasons in addition tothe rebound effect for this. First, most con-sumers who take advantage of these pro-grams are free riders; they would havebought those devices with or without thehandouts.164Second, the high initial cost ofmany of those technologies (like the fabled$40 light bulb) implies that it can take years

20

Legal or ders canreduce demand,but at best theyare much mor e

cost ly to developand enf or ce than

pr ice increases,whi ch l ead all

people, not justgovernment

lawyers, to th inkabout h ow to

conserve.


21/35

to recoup the costs of those investments,even after the subsidy. Other investments arefar more attractive by comparison.165Third,many of the wonder-techs are known to bepoor performers, trading off other consumerconveniences to eke out a little more efficien-cy at the margin. People are understandablyleery of pitching thousands of dollars worthof appliances with years of life left in them toembrace unknown technologies that oftenhave their own problems and save moneyonly after years of operation, if ever.

The hard truth is that people will conserveenergy only when they think that the incon-veniences of doing so are outweighed by themoney saved by such investments or by suchbehavioral changes. Reducing the marginal

cost of energy consumption is exactly thewrong approach. And even were it the rightapproach, government programs that puttheir faith in subsidized energy conservationrather than free energy prices helped give usthe California mess in the first place. Only byletting prices do their job can energy conser-vation be achieved. Everything else is politicalsmoke and mirrors.

Public OwnershipSome people in California argue for pub-

lic ownership in addition to price controlsand legal orders. They point out that the LosAngeles Department of Water and Powercontinues to sell power to ratepayers at pricesfar below wholesale prices in the Californiaspot market without incurring the debtsincurred by the investor-owned util it ies.

The correct framework in which to ana-lyze LADWP is the one we used in our dis-cussion of installed-capacity regimes.Currently, LADWP has excess hydro- andcoal-fired generators under contract. It sells

the excess on the spot market, earning largeprofits just like private owners of similarfacilities, and uses the profits to subsidize theconsumption of its current customersinstead of returning them to taxpayers as pri -vate utilities would return profits to share-holders.166

The LADWP policy has the appearance of

a free lunch, but it is not. Until the spring of2000, when those excess supplies becamevaluable on the spot market, in order to payfor the unused excess capacity, LADWP cus-tomers had to pay more for electricity thanthey would have paid in a spot market.

In addition, if LADWP supplements thosepower supplies with gasfired electricity, itsincremental costs are similar to those of pri-vate generators. But unlike the deregulatedwholesale market, LADWP prices its outputon a weighted-average basis passing on thelow costs of coal and hydro to consumersrather than charging for all output on thebasis of the price of the most expensive input,natural gas. As we demonstrated earlier, suchpricing is unsustainable because it encour-

ages extra demand, the costs of which aremore than incremental revenues. Eventually,LADWP will have to obtain more capacitythan it would under marginal-cost pricingbecause consumers are not charged enoughto reflect their incremental demands on thesystem and thus LADWPs ultimate costs arehigher than they otherwise would be becauseof its pricing policies.

Long-Term ContractsIn early February, the California legislature

adopted A.B. 1X, authorizing the state to direct-ly contract with generators and marketers forelectricity. Governor Davis, the author of theplan, hoped that 95 percent of the states elec-tricity demand could be secured through suchcontracts, leaving only 5 percent of the statesneeds to the spot market. The governor furtherhoped that 3- to 10-year contracts would allowCalifornia to lock in prices at a weighted averageof 5.5 cents per kWh ($55 per MWh). Davissexpectations were dashed, however, when 38power suppliers submitted bids at a weighted

average of 6.9 cents per kWh.167 Although thegovernor announced that the state had agreedto pay $42 billion for this power over 10 years,little else is known of the arrangements.168

But will the contracts really lower pricesfrom current spot levels? First, the 6.9 cent fig-ure is for nonpeak rather than peak supply.While information is sketchy, reports suggest

21

Only by lett i ngpr ices do t hei rjob can energyconservat ion be

achieved.Everyth ing else ipolitical smokeand m ir rors.


22/35

that bids for long-term peak supply averagearound 25 cents per kWh.169 Second, the 6.9cent figure is a weighted average. As discussedearlier, the proper price for electricity is not theaverage price but the price paid for the mostexpensive source used during any one-hourperiod. Even if the state were to resell thatpower to ratepayers at cost, it would presum-ably be sold on the basis of the average acqui-sition cost and thus would be priced too cheaprelative to an unregulated market.

If long-term contracts could hold downmarket-clearing prices in free markets, whatare we to make of the natural gas market?According to the American Gas Association,only 9 percent of all wholesale natural gas pur-chases are made from the spot market,170 but

the price spike occurred. If long-term con-tracts could dampen wholesale prices duringspikes, we wouldnt have seen anything likethe natural gas price explosion of 200001.

Theres also good reason to think thatlong-term contracts signed by the state willbe more expensive than long-term contractssigned by private parties. Because futurepoliticians will be tempted to renege on thosecontracts if spot prices fall below the contractprice, producers will include a premium intheir bids to cover that risk. Confidence in

the states good faith is not enhanced by thecurrent talk of confiscating plants and crim-inalizing pricing activity when prices are arbi-trarily thought to be too high.171

Regardless of the details of Californiasarrangements, the argument tha

california’s electricity crisis what

Documents