Assessing consumer benefits in the Ontario residential retail natural gas market: ontario residential retail natural gas market: Why marketer entry did not help

Energy Policy 109 (2017) 555–564

Contents lists available at ScienceDirect

Energy Policy journal homepage: www.elsevier.com/locate/enpol

Assessing consumer benefits in the Ontario residential retail natural gas market: ontario residential retail natural gas market: Why marketer entry did not help

MARK

Barb Bloemhof DeGroote School of Business, McMaster University, 1280 Main Street West, Hamilton, ON, Canada L8S 4L8

A R T I C L E I N F O

A BS T RAC T

Jel code: L51 Q48

When the institutions governing transactions in the energy sector changed in Canada thirty years ago, the changes were heralded as pro-competitive for natural gas markets, and subsequently authorities and some stakeholders have suggested that households would benefit from gas commodity marketer entry. Using an analysis of the institutions together with a model of the cost pass-through regulatory process, I show that households as a group would not benefit from purchasing the gas commodity from unregulated commodity marketers rather than buying it bundled with delivery from the regulated utility. I next use monthly price data collected from the public web pages of these unregulated sellers to confirm the theoretical prediction. Depending on the time period, five-year natural gas contract commodity prices average more than 75% higher than the utility's regulated cost of gas. On average over the sample period, signing up meant paying over C$400 more for the gas commodity annually. The analysis suggests that claims about the effects of policy change should be informed by a careful assessment of the specific institutional arrangements and the new incentives created by institutional evolution.

Keywords: Deregulation Institutions Natural gas Retail markets Unbundling

1. Introduction The idea that consumers benefit from a competitive energy market is uncontroversial (Fabrizio et al., 2007; Mansell and Church, 1995, p. 20; Ontario Ministry of Environment and Energy, 1996; Joskow, 2008). Economic theory predicts that, under certain conditions, the discipline exerted on the market by many sellers and buyers pursuing self-interested choices creates opportunities for improved efficiency (Tirole, 2003, p. 212).1 On the other hand, regulation creates significant compliance costs and dulls the incentives for efficient resource use and innovative (Rose, 2014, p. 2). Competition is viewed as the best way to get the most out of current and future economic opportunities (Crampton, 2003, p. 3). Moving toward more competitive energy markets was an important Canadian policy direction in the 1980s and 1990s (Doern, 2005, p. 8). Cudahy (2001, p. 155) argues that during that period the deregulation movement was “almost the signature cause” for American policymakers as well. Facilitating competition in the sale of gas to users is listed right

above protecting the public interest in the Ontario Energy Board's stewardship goals for energy markets under its regulatory purview in Canada's largest domestic natural gas-consuming province (Ontario Energy Board, 2014, p. 7).2 The empirical evidence about competitive energy markets has been mixed. Arano and Velikova (2009) find a more competitive natural gas market conferred benefits for residential consumers after industry restructuring in the United States. Brau et al. (2010) are much more circumspect on the experience of European Union consumers postreform. Spence (2008) finds that prices are systematically higher with competition in electricity markets, while Kwoka (2006) provides theoretical conditions under which the excess costs of electricity market regulation could dominate subadditivity in the underlying production function and therefore justify duplicating sunk infrastructure. Kwoka also finds that in his sample public ownership enhances efficiency. Cudahy (2001) discusses the incentives for surplus capture that deregulation creates, including how electricity markets in California were gamed by sellers. Rose (2001, p. 1296) suggests that

E-mail address: [email protected]. Efficiency is one of two criteria associated with using resources in ways that society values the most. The second criteria, equitable distribution of surplus in the economy, is normative, involving judgments of which agents in the market should have the surplus created by the resources. By contrast, efficiency is about maximal surplus from a given economic endowment of tastes, technology, resources and institutional rules. 2 The regulator's stewardship goals have evolved over its history, particularly after the early 1990s. The regulator's enabling legislation does not provide guidance about which of these goals should have priority. I am indebted to an anonymous referee for both of these observations. 1

http://dx.doi.org/10.1016/j.enpol.2017.07.038 Received 3 May 2017; Received in revised form 14 July 2017; Accepted 17 July 2017 0301-4215/ © 2017 Elsevier Ltd. All rights reserved.

Energy Policy 109 (2017) 555–564

B. Bloemhof

model's theoretical prediction: that, contrary to the claim of those who advocated for the policy change, the entry of new unregulated natural gas sellers has not conferred benefits to the average consumer in the form of lower prices in the Ontario residential retail natural gas market.

the gains from technical change that justified de-integrating some vertically-integrated natural monopolies had not yet been fully realized. A few papers look at the impacts of unbundling the natural gas commodity from naturally monopolistic pipeline services. Unbundled commodity sales are said to foster efficiency through entry and competition (Barcella, 1996; Hlásny, 2011, p. 202) but this assertion depends on the specific institutional arrangements that govern how energy is bought and sold in a particular market. The study by Lee et al. (2004) suggests that the impact of a change in rules on consumer surplus is also an empirical question. Casarin (2007) and Arano and Blair (2008) argue empirically that competitive supply after unbundling can confer benefits. Neither of these empirical papers discuss the implicit limits on commodity pricing power provided by the institutions per se. Unbundling and new seller entry in the Ontario residential market had unexpected and negative consequences for households. The trading rule changes initiated by the federal government in the upstream or wholesale market, after over a decade of price administration, created an environment where market forces could determine prices (the Hallowe’en Agreement, Canada, 1985). Consumers benefitted as market supply and market demand equilibrated at a lower wholesale price, removing a significant wedge between producers’ price and cost. However, unbundling interprovincial transportation services from gas commodity purchase made space for independent unregulated sellers to enter and market the gas commodity to households at retail prices, while the regulated distribution utility would continue to supply the gas commodity at its wholesale cost to households who did not sign up with a new seller. New electronic purchasing technologies and the new North American energy futures trading market opened in 1992 further lowered the cost of sourcing and purchasing the commodity on behalf of households, making entry into the unregulated sector cheap. A natural experiment in commodity pricing has been going on for thirty years. A market's institutional arrangements are critically important to its performance. The institutional evolution of the natural gas sector provides valuable insights about pitfalls that can be avoided in market reforms. Inferences about competition among retail marketers as a class, and the actual distribution of surplus in the unregulated sector, suffer from incomplete data.3 However, it is simple to model the distribution of surplus implied by the institutional arrangements before and after the change, and use marketers’ own historical unregulated prices to show that surplus was likely transferred to newly entered natural gas marketers and from consumers actually buying in the unregulated market.4 The first task is to analyze the institutional arrangements in the market. I describe what the federal government changed about how Alberta natural gas is sold and how those institutions were accommodated in retail markets. I next identify a key source of efficiency that works in the public interest that was not eliminated post-1985, and outline how the new rules created an opportunity for sellers to establish and exploit information asymmetry over the public in the transition to an integrated North American wholesale natural gas market. Then I present a model of the cost of getting natural gas from the Alberta wellhead to the burner tip in Ontario in the regulated and unregulated market segments, demonstrating the cost pass-through institutional arrangement of the regulated utility. Finally, using fixed price commodity contract data collected from a commercial website I confirm the

2. Formal institutional arrangements in the residential retail natural gas market Institutions, the laws, rules or habituated behaviours that accompany and facilitate exchange in a market, define the rights and possible actions available to agents in the market; they delimit both the potential surplus available in a trade and the distribution of that surplus between buyers and sellers (North, 1990, p. 4; Loehman and Kilgour, 1998; Ostrom, 1986). Institutional arrangements are the (often enforceable) rules of the trading game. Formal institutional arrangements are the laws, regulations and government policies that shape and constrain economic transactions and behavioral choices available to agents in those markets, while corporate policies and habituated ways of transacting are examples of informal institutional arrangements. Consider, for example, the institutional arrangements that give a private pipeline company the right to provide natural gas carriage in a particular defined franchise area. The constitution gives the government authority to make rules within its jurisdiction. The government in term creates a regulator endowed with a legally defined and limited authority to make rules about entry and operations in its jurisdiction.5 The regulator vets pipeline proposals and confers to a single pipeline the right to operate in a particular franchise area, in the public interest. A monopoly structure clarifies what pipeline is unambiguously responsible for the safe movement of natural gas in that area, and incents a more efficient investment in infrastructure characterized by subadditive cost.6 In this way, the institutional arrangements set out how the rules of commerce operate for all potential and actual entrants into the pipeline carriage enterprise. Institutions delimit the size of the potential surplus available in a trade, and define the means to claim or defend a particular distribution of this surplus between buyers and sellers, so institutional differences can explain significant differences in prosperity between otherwise similar economies (Feeny, 1988, p. 159; North, 1994, pp. 361–362; Rose, 2001, p. 12961). They change relatively slowly and are typically known or knowable. A new set of institutions may create a larger pool of gains from trade or make such gains easier for one side of the trade to capture. Anticipated or perceived incremental gains from trade thus provide an incentive to make or lobby for institutional change. There is no presumption that such institutional evolution is socially efficient (North, 1994, pp. 61–62); presumably, though, those with the power to effect institutional evolution stand to benefit from it. The total surplus available to split between sellers and buyers institutional change could 5 In Canada, the respective provincial pipeline regulators are: the British Columbia Oil and Gas Commission, the Alberta Energy Regulator, Saskatchewan Ministry of Energy and Resources, the Manitoba Public Utilities Board, the Ontario Energy Board, Régie de l′énergie du Québec, the Newfoundland and Labrador Board of Commissioners of Public Utilities, the New Brunswick Energy and Utilities Board, the Nova Scotia Utility and Review Board, the Yukon Environmental and Socio-Economic Assessment Board, the Northwest Territories Public Utilities Board and the Nunavut Impact Review Board. The federal interprovincial pipeline regulator is the National Energy Board, with some aspects of safety regulated by the Transportation Safety Board. Notice that only one level of government regulates a particular aspect of operations: here too economies of scale and jurisdictional clarity justifies avoiding duplication. If the pipeline crosses provincial or international boundaries, then the federal government sets the regulation; if the pipeline crosses international boundaries, then authorities from both countries must approve the pipeline. 6 A single large pipe confers significant volumetric economies of scale. Duplicating the pipeline's sunk investment (in pipe and compressor network infrastructure, including engineering and construction) and ongoing business expenses is not an efficient use of society's scarce resources. See Mansell and Church (1995, p.7).

3 The definitive way to examine the impact of the 1985 institutional changes is to map where the difference between the price and cost of natural gas went: either to fixed costs or to surplus rents. None of the independent sellers have agreed to share proprietary cost data, and a corollary of my argument is that they have good reason to protect this information. Of course, the data in the public domain is sufficient to investigate the impact of the new rules. 4 The opportunity to choose is itself argued to enhance consumer welfare, although this argument is harder to make with an homogenous product like natural gas.

556

Energy Policy 109 (2017) 555–564

B. Bloemhof

Fig. 1. Average Alberta Market Price (current dollars per GJ). Source: Alberta Energy (2017).

not actuarially-quantifiable risks, notwithstanding active futures markets, and the price in the spot market could go up or down (a two-sided risk). Nonetheless, consumers were well-used to the concept of insurance, and anecdotal evidence suggests that a number of people signed up with marketers on the basis of that pitch. 7 In the regulated part of the market, commodity prices by each of three privately-owned distribution utilities are reviewed by the Ontario Energy Board in the public interest.8 The utility provides local gas carriage and storage services, as well as billing services, to any gasusing household or business within their own exclusive franchise area, whether the utility buys the commodity for them or not. Every quarter, the regulator examines each utility's pipeline cost of service and approves the prices charged for delivery services, limiting the amount of monopoly rent or surplus earned on pipeline asset services. The regulator also reviews the utility's projected and actual cost paid in the wholesale market for the gas commodity, and approves a retail price per cubic meter that just covers the commodity purchase as a disbursement without mark-up. Households who do not sign up with a marketer have their gas commodity supplied by the utility and pay the utility's actual gas acquisition cost, subject to routine quarterly adjustments up or down as the wholesale price fluctuates. Understanding this difference in institutional arrangements governing commodity pricing is crucial to the question of whether the entry of unregulated marketers was helpful to consumers. Before 1985, all distribution pipelines were merchants of delivered natural gas, combining their pipeline delivery services with Alberta gas purchased from the marketing department of the transportation pipeline (which was itself a gas merchant, buying from producers and selling to the utilities). The job of acquiring the gas fell to the party with the most knowledge of how to do this, and therefore the lowest cost. 9 The cost of

rise, fall or stay the same. Equally, sellers may be able to devise responses to rules that facilitate surplus capture from consumers, making the distribution of surplus less fair even if there has been no change in total surplus. Rulemaking itself is not costless and incomplete information further affects equity. Market participants intent on surplus capture may not want to disclose all the information they have about their situation or choices. After the federal government removed price controls on wholesale natural gas in 1985, a cashflow problem on the transportation pipeline serving Ontario created a way for a new class of market participants to capitalize on asymmetric information. The transportation pipelines split off their marketing activities as separate companies and offered contract carriage at the regulated rate to anyone buying gas. The largest industrial users immediately explored the opportunity to lower their energy costs right away through direct commodity purchase from the producer after a decade of unprecedented real energy prices. For smaller gas users, the benefits of buying their own gas commodity might not offset the expense. Unregulated marketers began to act as brokers by aggregating demand in the commercial and multi-residential sectors. By 1992, marketers began canvassing individual households in Ontario to sign them to fixed-price gas commodity contracts. Until then, residential consumers had only ever bought their gas from the regulated distribution pipeline in their jurisdiction. The Ontario Energy Board required these “merchant” pipelines to pass through to consumers the actual weighted average cost of the gas commodity, and regulated their cost of delivery services (cost of service regulation; see Mansell and Church, 1995, p. 57). In assessing the utilities’ earnings on infrastructure investment, the regulator had always distinguished between the cost of providing delivery services and the pass-through cost of the gas commodity, although the utilities billed consumers monthly for the commodity and the service as a bundle. If a household were to sign up with one of several unregulated gas marketers, the amount per cubic meter of gas consumed was not going to change over the life of the contract (typically one to five years). The regulated utilities would also undertake billing and collection for all marketers in their respective franchise areas, reflecting the commodity charges as a separate line on the household's monthly bill. Fig. 1 shows that the nominal wholesale price of natural gas can and does vary significantly over the short term. Fixed price natural gas contracts were sometimes marketed as insurance against the sort of rising energy prices that consumers had experienced over the 1970s and 1980s. Insurance protects the buyer from an actuarially-quantifiable risk, typically a one-sided loss which can be partially underwritten by group participation in risk pooling. Wholesale natural gas price movements are

7 A better explanation of the prospect than insurance is the “money pump” characterized by Rabin and Thaler (2001) because of the size of the premium, the absence of a one-sided loss and the information asymmetry. I am grateful to a conference participant for providing me with an opportunity to clarify the misapprehension when he indicated that, if a similar “insurance” were offered for retail gasoline market price movements, he would not sign up for a five-years’ contract for gasoline at a price fixed at 20–70% more than the current price posted at the gas pump. 8 The municipalities of Kingston and Kitchener own and operate their respective natural gas distribution networks. As municipal operations that do not sell carriage to customers outside of their own franchise area, neither is regulated by the Ontario Energy Board. 9 The distribution pipelines sourced gas from the (monopoly) merchant transportation pipeline to which they were intertied, but it was more than just historical fluke that originally endowed the pipelines with the property right of being merchants of gas. Initially a nuisance by-product of oil production in Essex County in southwestern Ontario, natural gas was a new product provided to Ontarions by integrated drilling,

557

Energy Policy 109 (2017) 555–564

B. Bloemhof

business. To see why, note that natural gas is a standardized homogenous and undifferentiated commodity,11 and consider for each of these types of providers in turn the relationship between their price and their cost of natural gas. Both purchase the commodity in a wellfunctioning wholesale market at a cost CC per cubic meter and arrange to have it shipped on an interprovincial transportation pipeline to the distribution utility junction at a cost of CT per unit. From there, the distribution utility charges all customers a regulated delivery cost of CD per unit regardless of commodity provider.12 Provided that both categories of buyer are sophisticated traders, neither has a bargaining disadvantage in the wholesale market and both will pay CC for the commodity. Then the per-unit cost of delivered natural gas (CDNG) is the sum of the wholesale commodity cost and the charges for pipeline carriage services that take the commodity from the wellhead to the burner tip:

service regulatory process for the pipeline's product, which was delivered natural gas, considered the reasonable and actual cost of gas as an input, and approved a cost of service that implied no more than the allowed rate of return on the approved capital stock (Mansell and Church, 1995, pp. 59–61). The Ontario Energy Board Act did not change when the federal government began administering the price of natural gas in 1974, or later when it dismantled that price administration regime in 1985.10 In testimony to the federal Standing Senate Committee on Energy and Natural Resources (Canada. Senate Standing Committee on Energy and Natural Resources, 1987, pp. 81–82), the chair of the Ontario Energy Board confirmed that the regulator could treat carriage and the commodity as separate charges. The utility had always presented and defended before the regulator its input cost of natural gas. Clearly, the pipeline infrastructure made sense as a rate base for pricing delivery services regardless of who owned the gas in transit. There was thus never a regulatory barrier to an unregulated demand aggregator entering the market to buy gas in Alberta on behalf of customers in Ontario. This regulatory approach also ensured that households paid the lowest possible cost and enjoyed the highest possible surplus for the commodity, both before and after 1985. The cost of the gas commodity is treated as an ex post knowable disbursement that is unrelated to the pipeline asset when calculating just and reasonable charges for gas delivery services. The regulator updates the weighted average cost of the natural gas commodity quarterly as more is known about its actual cost. Any policy change, but particularly one as significant as decontrolling producer prices, should be reviewed in light of the extant institutional arrangements because they matter to the outcome (Rose, 2014, pp. 12–14; Tirole, 2015, p. 1666). The Hallowe’en Agreement certainly did benefit consumers by removing the wedge between the price paid to producers for gas and its resource cost, and fostering a “flexible and market-oriented pricing regime” for the gas commodity (Canada. Department of Energy, Mines and Resources, 1985, p. 1) that lowered its wholesale price continent wide. With its increasing returns to scale on carriage services and economies of scope on storage, load balancing and billing, the pipeline sector remains monopolized and regulated in the public interest. Regulated distribution utilities have always been restricted to passing through to their customers only the reasonable and actual wholesale cost of natural gas. But the formal institutional rules enabled surplus transfer from households to marketers. Regardless of whether Ontario's unregulated commodity market was competitive as asserted by the regulator (Ontario Energy Board, 2015a), marketers must mark up their sales in order to cover their costs of business, even before they earn any profits.

CC + CT + CD = CDNG . This cost is exactly what consumers pay if their gas is purchased for them by the regulated distribution utility. The utility passes what it paid for the per-unit cost of natural gas (CC) and its transportation (CT), directly on to its gas commodity customers on whose behalf these expenses were incurred.13 The distribution utility covers its total costs and earns its return from the per-unit charge for delivery of all gas on its own network (CD). Fixed pipeline, overhead and staffing costs, and variable pipeline operating costs, as well as a regulator-approved return on investment, are included in calculating CD; furthermore, CD is calculated on the knowledge that it will be charged on all volumes delivered, whether they are purchased in the wholesale market by the distribution utility or by one of the natural gas marketers for fixed-price contract customers. The regulatory process allows only defensible actual expenses for the gas commodity and its transportation to be charged to the utility's commodity customers as the per-unit cost of delivered natural gas (CDNG); the regulator prevents the utility from charging any markup on the natural gas commodity or what is paid to ship the gas to the household. In contrast, for unregulated marketers, the contractually fixed gas commodity price charged to their customers must be inflated by some positive margin to cover fixed and variable overhead costs of its operations, even before any profit or return to entrepreneurship is included. The marketer's contractually fixed unregulated natural gas supply price is:

PM = (CC + margin) Gas purchased by a marketer also must be transported interprovincially from the gas field, and the marketer's contract specifies that the fixed-price contract customer will be charged for transportation. The regulated transportation pipeline charges the same amount per unit (CT) on all volumes carried regardless of which commodity provider purchased the gas.14 As the provincial pipeline regulator requires, the

3. A Model of natural gas cost and prices

11 Only the institutional arrangements that form the terms of sale differentiate the natural gas commodity from different sellers. Because some naturally-occurring impurities are corrosive, or become noxious or carcinogenic when burned, natural gas is refined to pipeline-quality dry gas purity before shipping, to avoid damaging the delivery infrastructure (collection, transportation, and distribution pipelines as well as gas-using appliances) and to form predictable and safe products of combustion when burned (water and carbon dioxide; a small amount of mercaptans is added to make unburned natural gas detectable by its smell). 12 Without loss of generality, the distribution charge can include a fixed component for monthly billing and a variable amount for storage and load balancing. 13 The utility updates quarterly the weighted average cost of natural gas purchased on behalf of its customers as the unpredictable, weather-sensitive demand pattern of consumption becomes known in hindsight. Transportation charges also change less frequently according to the regulatory decisions of the interprovincial pipeline regulator, the National Energy Board. 14 Matching the distribution utility's transportation charge is a recent phenomenon. Distribution pipelines in some franchise areas offer space contracted for their customers to unregulated marketers at cost. In those instances there is no difference between transportation charges for regulated and unregulated gas commodity providers, and the

Households buying the gas commodity from unregulated marketers pay higher per unit natural gas costs compared to households buying the commodity from utilities. The reason is that unregulated marketers must charge a higher commodity price than regulated utilities on average because they earn returns on different aspects of the gas

(footnote continued) transportation and distribution companies that had to be contained in pipes to be sold; the pipes were initially regulated for safety by the Ontario Fuel Board. Although Essex gas was Canada's first energy exports (to Buffalo in 1891 and Detroit in 1895; see Stenson (1985, p. 4), reserves in this geological formation served a number of communities from Sarnia to Guelph, supplied using what was for the time cutting edge pipeline technology. Toronto had to wait until 1955 to move onto natural gas at a third to a quarter of the price of manufactured gas (Jones, 1955). 10 The 1980 and 1990 consolidations of the Ontario Energy Board Act (Ontario, 1998) are substantively the same; the only changes enhanced clarity (“notwithstanding” changed to “despite”) or expanded gender-neutrality.

558

Energy Policy 109 (2017) 555–564

B. Bloemhof

marketers must charge more per cubic meter for the gas commodity than the actual wholesale gas cost, at the very least by enough to cover their reasonable and actual business overheads. Put another way, only if their operational costs and profit expectations are zero could marketers offer similar commodity costs that regulated utilities charge to customers. Given the institutional arrangements, only if marketers systematically lost money could they provide the commodity at a lower cost than the utility.

delivery charge (CD) also applies to all volumes carried on the distribution utility's system. Then the household's cost per unit of delivered natural gas under a fixed-price contract is the sum of the marketer's commodity price and the charges for transportation and distribution services:

PM + CT + CD = (CC + margin) + CT + CD = CDNG + margin. The cost of delivered gas from an unregulated marketer is higher than the distribution utility's delivered natural gas cost (CDNG) by the positive amount of the margin. As long as the marketer has any fixed costs including a return to entrepreneurship, the marketer cannot price lower than the distribution utility's delivered natural gas cost (CDNG). The consumer will pay less to just let the utility source and supply their gas than to sign a commodity contract with a marketer. In a dynamic model, the marketer can offer new customers a lower price than it had offered in the past. This possibility arises because the term of a fixed price contract ranges from one to five years, a span of time that can encompass significant gas cost variability. If the wholesale cost of natural gas is falling relative to recent historical levels, the marketer can price discriminate and subsidize the losses made on some contracts by higher gains on existing contracts. In fact, it is precisely the elimination of commodity cost variability that marketers are promoting. However, the sum of consumer surplus across all fixedprice contract terms in the marketer's portfolio must be strictly less than the sum of consumer surplus in the counterfactual case of utility supply, where the actual wholesale commodity cost is passed on directly to consumers without any mark-up to cover the marketer's costs or economic profit. The analysis is conservative in that the model assumes (1) the interprovincial pipeline charges marketers and utilities the same regulated transportation cost CT, and (2) in the wholesale market, utilities do not have a bargaining advantage over marketers. Whether or not the transportation charge is the same as for utilities is proprietary information; however, the unregulated marketers can certainly pass on to their consumers a higher transportation charge.15 In contrast, the regulated distribution utility is required by the institutional arrangements to simply pass on the transportation charge without markup. Regarding the efficiency of the wholesale market, there is some scope for price discrimination. Natural gas producers may offer volume discounts or charge a premium for a particular delivery to a buyer which is perceived as having questionable creditworthiness. Both of these factors favour the regulated utilities, which have higher volumes than any individual marketers and own durable pipeline assets that imply longevity and underscore a secure cash flow.16 Furthermore, volume discounts and bargaining advantages accruing to the distribution utilities would be passed on to consumers through the regulatory institutional arrangements. The entry of natural gas marketers could not have improved commodity prices for consumers. The regulation process has always ensured that households that bought their gas commodity passively from the distribution pipeline utility paid the lowest possible gas commodity cost. Because they have no other product, natural gas

4. Long term fixed price contracts offered by natural gas marketers: The data The typical consumer may not be aware of the institutional rules or the marketers’ pricing implied by those rules illustrated in the model above. However, the marketers’ own prices over the past thirty years provide some empirical evidence to refute the claim that unregulated sellers can provide the commodity at less than the regulated distribution utility's cost. This paper contributes a consistent set of monthly unregulated natural gas price offers for the Ontario residential retail sector over the period from 2005 to December 2016, and selected data from the late 1990s to 2005. I focus here on the most recent data to remove any temporary influence of minor rules changes in the Ontario market before 2002. These data from the new market institutional arrangements provide a straightforward confirmation of the conclusions from the institutional analysis presented above. Fig. 2 shows the utilities’ commodity charges passed through to consumers. Table 1 reports the average unit cost of natural gas provided by Ontario regulated distribution utilities. Natural Resources Gas has a higher regulator-approved cost of commodity supply (still passed through to customers without markup) because it does not own any storage facilities and instead purchases storage services from Union Gas. With that adjustment, the regulated distribution utilities pass through similar per-unit amounts for the gas they source for their customers (CDNG). In contrast, Fig. 3 shows considerable variability between marketers’ commodity prices on similar term contracts. The highest price offered for a five-year term is nearly 50% higher than the lowest price. Individual marketers often list more than one contract length, but their prices for other contract lengths seldom vary by more than a few cents. The regulated rate for supply from the utility in December 2010 was 15.4 cents per cubic meter before gas cost adjustments,17 or between 20% and 55% lower than the unregulated price offers, as predicted. The Transportation Note in Fig. 3 is also interesting, because the buyer of the gas also arranges for transportation on the ex-Alberta pipeline. In this example, eight out of ten marketers charged more for transportation (CT) than the rate passed through by the regulated utility. The differential could exceed the price advantage to a longer contract term. Figs. 4 and 5 show the lowest five-year fixed-price contract available in each month against the utility's cost of gas over the period since January 2005.18 Marketers enter and exit frequently, sometimes by takeover and reorganization of existing marketers.19 With rare and short-duration exceptions, the lower envelope of marketers’ commodity price lies above the average cost of natural gas that is charged to households supplied by the regulated utility.

(footnote continued) model is exact. However, historically, the utility's and marketers’ transportation charges to consumers have been very different. I return to transportation charge differences below. 15 Before 2010 the typical marketer's retail contract specified a higher charge for transportation than that paid by the distribution utility; depending on the seller, this charge was between 1.8 and 2.7 times the transportation charge paid by utility commodity customers. Only the individual marketer knows whether this practice was because the marketer genuinely faced higher transportation rates, or because the marketer was also marking up transportation purchased on behalf of its consumer. 16 The historical data that I have collected show that marketers enter and exit the market without warning. I am indebted to Peter Vilks for pointing out to me that a credible cash flow may cause gas producers to view utilities more advantageously than marketers.

17 Gas cost adjustments can either reduce or increase the gas commodity cost, but more often they reduce it and have historically been much larger in absolute value when they reduce the gas cost. In the case illustrated in Fig. 3, the net regulated price of the natural gas commodity was 13.8 cents per cubic meter after a price adjustment of −1.64 cents. 18 Depending on the marketer, the five-year term could be higher or lower than the other fixed-price contract terms offered in a particular month. The five-year contract term is the most commonly offered contract term. The Ontario Energy Board focused on the five-year contract term in its Report to the Minister (Ontario, 2015b:15). 19 The number of marketers operating in any jurisdiction has varied from one to twelve over the time period from January 2005 to December 2016, and most marketers were not present continuously in any jurisdiction over that time.

559

Energy Policy 109 (2017) 555–564

B. Bloemhof

Fig. 2. Regulated distribution companies' regulator-approved rates (current cents per m3). Source: Ontario Energy Board (2017).

To see how much more a household on a fixed-price contract paid for the commodity, consider the natural experiment shown in Figs. 6 and 7, where the lowest unregulated commodity price from a marketer each month is compared to the average of the next sixty months’ actual cost of natural gas from the regulated utility (the comparable actual commodity cost that obtained over the contract term). Each observation shows how much more a household signing up with a marketer in that month paid over the next five years for the gas commodity, before considering any differences in transportation charges. Over a typical year, natural gas from a marketer cost about $34 more than it cost when passively purchased from the regulated utility. Although the gap between lowest unregulated price offer and the regulated commodity cost has recently been narrowing, the entry of marketers has led to substantial consumer surplus loss for anyone who signed up for fixed price commodity supply, peaking in winter 2009 at more than two and a half times the utility's gas cost. These impacts are summarized in Table 2. The data show that even buying from the lowest-priced marketer could not be better for a household than remaining on the regulated utility's commodity supply. Converting this finding into a measure of the total surplus transferred from consumers to unregulated marketers over the study period would require sales volume data for each natural gas marketer that is not available in the public domain. It is reasonable to conjecture that the different marketers have diverse and timevarying sales volumes and market shares, but without that proprietary quantity data the actual total surplus loss to consumers over the study period is unknowable.

Table 1 Average regulated residential natural gas commodity rates (cents per m3). Source: Ontario Energy Board (2017). Year

Union Gas

Enbridge

Natural Resources Gasa

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

29.22 35.16 25.79 30.52 19.93 13.04 13.33 9.43 10.93 19.37 13.09 9.83

27.70 31.32 27.96 31.13 18.55 16.01 12.34 9.12 11.35 17.21 15.15 10.84

– 42.68 35.57 37.92 31.16 29.39 20.87 18.67 19.39 27.45 21.53 16.50

a Natural Resources Gas rates include transportation charges of a similar order to those paid by Union Gas and Enbridge, adjusted for distance; and storage charges paid to Union Gas. Natural Resources Gas is the third Ontario Energy Board regulated distribution pipeline, serving some 8500 customers in and around Aylmer (source: conversation with Stacey Ricketts, Manager, NR Gas, 5 July 2016).

Over the study's time period, the majority of marketers’ price offers were above, sometimes substantially above, the weighted average cost of gas passed through to and paid by system gas consumers who had simply not signed up (or declined to sign up) with a natural gas marketer. The lowest per-unit commodity price offered by a natural gas marketer was lower than the commodity cost passed through by a regulated utility in only 8.3% of the months (Enbridge area) and 13% of the months (Union Gas area).20 These commodity prices also exclude the cost of shipping gas on the transportation pipeline. Currently, most marketers match the local utility's transportation charge. Historically, marketers’ transportation charge exceeded the respective utilities’ transportation charge by as much as 150%. Whether the higher price was due to an actual cost difference, or due to opportunism, is proprietary to the marketers. These data show that only rarely is it possible to sign up for unregulated gas at a price below the regulated commodity cost. For every person that signs such an advantageous contract, the marketer must have other consumers signed up for long-term supply at above cost to cover the marketer's administration, overhead, and economic return to investors.

4.1. Other sources of consumer value Although it is a homogenous product, there are two ways to add economic value to the natural gas commodity. Both are related to the dramatic variability of the household's total expenditure on natural gas across the seasons. Fig. 8 shows the variability due to weather-driven demand for heating over the course of a year (the load shape) in a southwestern Ontario household. The dominant driver of this consumption pattern is the space heating load; all households with natural gas furnaces will have a similar pattern in their load shape reflecting between five and seven times more energy consumed in the winter months than in summer. These seasonal changes in demand are far more significant sources of monthly variability in expenditure than marginal changes in average unit commodity cost. Seasonal expenditure variability puts a claim to surplus into the public domain (Barzel, 1997). The first source of consumer value is a predictable bill each month. There is a class of agents to the transaction that can provide levelized monthly billing, based on a projection of annual consumption, at a lower cost than what the consumer is willing

20 The Enbridge area data are commodity cost inclusive of transportation (CC + CT) for nearly two-fifths of the months in the comparison. In either jurisdiction, unregulated marketers could post a much higher charge than the regulated utility passed through for the same transportation service. For example, in the Union Southern Ontario area, transportation charges are 37–65% higher than the regulated utility's transportation cost; Enbridge area marketers were not even posting transportation rates at that time.

560

Energy Policy 109 (2017) 555–564

B. Bloemhof

Fig. 3. Screenshot of marketer posted price offers. Source: M3 & W website (10 December 2010). Used with permission.

to pay for it predictability.21 Unfortunately, that class of agents is not the marketers. All regulated utilities have been providing this valuecreating institutional arrangement at no additional charge since well before marketers entered the sector, because it is congruent with their own value position: predictable bills may be easier for customers to

pay, which lowers arrears and collections costs for the utility; and levelized billing provides a small advance on the utility's receivables in the fall, when cash flow is being applied to gas purchased and stored for the winter.22 Because the regulated utilities have always collected household meter data, they have the cheapest cost of billing service provision and therefore the lowest cost of providing equal billing. Second, regulated gas consumers who are not on levelized billing gain value from the timing of quarterly revisions to the cost of gas,

21 The household's annual consumption from last year is scaled by the utility's heating degree days estimate (from their forecast model) to arrive at expected annual natural gas consumption for the upcoming year. That amount is divided by 12 and billed out each month until July irrespective of actual consumption. The bill for August reconciles the estimated and actual use for the year.

22 The cash flow motive may explain why levelized billing is available to all natural gas accounts, regardless of whether the utility or a marketer provides their commodity.

561

B. Bloemhof

Energy Policy 109 (2017) 555–564

Fig. 4. Comparison of lowest-priced gas marketer & regulated gas cost, current cents per m3, Union Gas distribution franchise area, 2005–2016; five-year contract term. Source: Ontario Energy Board (2017) and M3 & W (various dates).

Fig. 5. Comparison of lowest-priced gas marketer & system gas cost, current cents per m3, Enbridge Gas distribution franchise area, 2005–2016; five-year contract term. Source: Ontario Energy Board (2017) and M3 & W (various dates).

Fig. 6. Lowest posted unregulated and regulated commodity costs, Union Gas southwestern Ontario area (lowest 5-year fixed-price contract; c/m3).

Fig. 7. Lowest posted unregulated and regulated commodity costs, Enbridge area (lowest 5-year fixed-price contract; c/m3). 562

Energy Policy 109 (2017) 555–564

B. Bloemhof

of wholesale natural gas. This form of market organization is consistent with a need for marketers to protect themselves from volatile wholesale price movements. Without proprietary information on their costs, it is not possible to address the question of whether marketers price competitively, in the economic sense of the word. However, the degree of competitiveness in the unregulated market is not the point. The value of economic competition is the maximization of surplus. The institutional arrangements that compel the regulated utility to charge the wholesale cost of gas without markup already provide the consumer with the most surplus available from the purchase of wholesale natural gas. Marketers could do no better. A few of the arguments in this paper are echoed in a recent report by the regulator to the Ontario Minister of Energy (Ontario Energy Board, 2015b) but a stronger set of conclusions follow from reflecting fully on the nature and evolution of the market institutions and from the marketers’ price data presented here. The possibility of a fixed contract price benefiting consumers in a rising price environment (Ontario Energy Board, 2015b:16) rests on the marketer's ability to cross-subsidize losses on those contracts with profits earned on other contracts priced above the wholesale price paid for the commodity. Wholesale prices in this market move up and down; a fixed price only benefits the marketer in a falling market, and it is only protective to some consumers in a rising market. Allowing this form of commerce pits consumer against consumer in a new (for energy markets) and inequitable way, particularly when many of those on unregulated contract supply are apparently not aware of making that choice (Ontario Energy Board, 2015b:8). Per-unit price smoothing across time is not valuable in a product where weather-driven quantity of energy demanded makes total monthly expenditure on natural gas (number of units times per-unit price) highly variable, and where that monthly expenditure variability has already been addressed by the distribution utility's equalized billing program. Finally, it is not clear whether the claim that some benefits occurring during the early period of rising prices (Ontario Energy Board, 2015b:15) would stand up once the higher prices for transportation on unregulated gas are taken into account.

which dampens the variability of wholesale price movements. Winter increases in the wholesale price of natural gas tend to be incorporated into the second quarter revision, after temperatures have warmed up and household demand for heating has fallen. The regulated price tends to be lower during the prime heating months, dampened either by the regulatory lag time needed to document gas costs or the flush state of distribution companies’ storage facilities. Of course, all cost changes will be passed through in the long run, but the timing of those revisions works countercyclically to weather-related demand, smoothing total natural gas expenditures to the benefit of the householder. Marketers may attempt to differentiate their product by offering incentive features like furnace servicing and Airmiles© bonus points, presumably to enhance market share. These incentive features do not fall from marketers’ own production space; they are not transactional assets that a marketer can leverage as the residual claimant of the surplus in the public domain (Barzel, 1997:8). These incentives are not without value, however presumably they are entirely financed by the consumer herself in the form of a higher commodity price, unlike the value or surplus due to levelized billing. Historically, in most time periods and utility franchise areas, there were only a small number of natural gas marketers whose prices changed infrequently and by small amounts relative to the movement

Table 2 Summary of Natural Gas Commodity Cost Comparison (cents per cubic meter). Union Gas southwestern Ontario

Enbridge greater Toronto

Average Commodity Cost, full sample (January 2005–December 2016) Lowest unregulated price each 24.53 25.56 month (5-year contract term) Actual regulated cost each 19.14 19.06 month Difference 28.2% 34.1% Average commodity cost in natural experiment (January 2005–April 2012) Average of lowest unregulated 30.54 30.86 price,5-year term Comparable regulated average 17.37 17.42 cost over the next 5 years Difference 75.9% 77.2% Annual household commodity expenditure in natural experiment (January 2005 – a April 2012) Lowest unregulated prices $ 946.74 $ 956.66 Regulated cost $ 538.47 $ 540.02 Difference $ 408.27 $ 416.64 (75.8%) (77.2%) Billing difference per month $ 34.02 $ 34.72 ($C)

5. Conclusions and policy implications The Ontario residential retail market provides an environment in which to consider the public interest impact of unbundling natural gas and delivery. The institutional analysis shows that dismantling over a decade of Canadian federal and producing province energy price administration in 1985 certainly provided benefits to all natural gas users through lower wholesale prices. However, the residential consumer did not benefit from the attempt to contrive a competitive retail commodity market by unbundling commodity and delivery and facilitating new seller entry. The regulator's longstanding practice of

Notes: Longest possible counterfactual for a 5-year contract. a Annual residential natural gas consumption of 3100 m3 is M3 & W's working assumption for comparison purposes in their marketing materials.

Fig. 8. Annual Load Shape for a southwestern Ontario household (m3). Source: author’s bills.

563

Energy Policy 109 (2017) 555–564

B. Bloemhof

news release). 31 October. Energy, Mines and Resources Canada, Ottawa. Canada. Senate Standing Committee on Energy and Natural Resources, 1987. Fourth Proceedings on: Examination of the Production and Use of Coal in Canada. Issue No. 10. 8 June 1987. Queen’s Printer for Canada, Ottawa. Casarin, A., 2007. Efficient industry configurations in downstream gas markets: An empirical assessment. Energy Econ. 29, 312–328. Crampton, P., 2003. Competition and efficiency as organising principles for all economic and regulatory policymaking. Prepared for Proceedings of the First Meeting of the Latin American Competition Forum (7-8 April). OECD/IADB, Paris. Cudahy, R.D., 2001. Whither deregulation: A look at the portents. N. Y. U. Ann. Surv. Am. Law 58, 155–186. Doern, G.B., 2005. Canadian energy policy and the struggle for sustainable development: Political-economic context. In: Doern, G.B. (Ed.), Canadian Energy Policy and the Struggle for Sustainable Development.. University of Toronto Press, Toronto, 3–50. Fabrizio, K.R., Rose, N.L., Wolfram, C.D., 2007. Do markets reduce costs? Assessing the impact of regulatory restructuring on US electric generation efficiency. Am. Econ. Rev. 97, 1250–1277. Feeny, D., 1988. The demand for and supply of institutional arrangements. In: Ostrom, V., Feeny, D., Picht, H. (Eds.), Rethinking Institutional Analysis and Development: Some Issues, Choices, and Alternatives.. Institute for Contemporary Studies Press, San Francisco, 159–209. Hlásny, V., 2011. Economic determinants of deregulation in the gas distribution market. J. Econ. Policy Reform 14, 201–213. Jones, O., 1955. Natural gas comes to Toronto. Address to the Empire Club of Canada (17 March). 〈http://speeches.empireclub.org/60399/data〉. (Accessed 2 April 2016). Joskow, P.L., 2008. Lessons learned from electricity market liberalization. Energy J. 29 (special issue #2), 9–42. Kwoka, J.E., 2006. The role of competition in natural monopoly: Costs, public ownership, and regulation. Rev. Ind. Organ. 29, 127–147. Lee, J.-D., Lee, J., Kim, T.-Y., 2004. Ex-ante analysis of welfare change for a liberalization of the natural gas market. Energy Econ. 26, 447–461. Loehman, E.T., Kilgour, D.M., 1998. Introduction: Social design for environmental and resource management. In: Loehman, E.T., Kilgour, D.M. (Eds.), Designing Institutions for Environmental and Resource Management. Edward Elgar, Cheltenham, UK (Northampton, MA), 1–25. M3 and W, Natural gas choices for your home, Energyshop, 2017. 〈http://www. energyshop.com/es/homes/gas/gas-prices.cfm〉 (Accessed monthly 2005 to 2017). Mansell, R.L., Church, J., 1995. Traditional and Incentive Regulation: Applications to Natural Gas Pipelines in Canada. The Van Horne Institute for International Transportation and Regulatory Affairs, Calgary. North, D.C., 1994. Economic performance through time. Am. Econ. Rev. 84, 359–368. North, D.C., 1990. Institutions, Institutional Change and Economic Performance. Cambridge University Press, Cambridge. Ostrom, E., 1986. An agenda for the study of institutions. Public Choice 48, 3–25. Ontario, Ontario Energy Board Act. Statutes of Ontario, 1998. ch. 15, Schedule B. (http://www.e-laws.gov.on.ca/html/statutes/english/elaws_statutes_98o15_e.htm) (Accessed 5 April 2011). Ontario Energy Board, 2014. Decision and interim order. EB-2014-0039 (27 March). 〈http://www.rds.ontarioenergyboard.ca/webdrawer/webdrawer.dll/webdrawer/ search/rec&sm_udf10=eb-2014-0039&sortd1=rs_dateregistered&rows=200〉 (Accessed 25 February 2016). Ontario Energy Board, 2015a. What the OEB does NOT regulate: Gas rate updates. 〈http://www.ontarioenergyboard.ca/OEB/Consumers/Natural+Gas/Natural+Gas +Rates〉. (Accessed 25 February 2016). Ontario Energy Board, 2015b. Consumers come first: A report of the Ontario Energy Board on the effectiveness of Part II of the Energy Consumer Protection Act, 2010 (28 May). 〈http://www.ontarioenergyboard.ca/oeb/_Documents/EB-2014-0158/ OEB_Report_to_Minister_ECPA_Review_20150528.pdf〉 (Accessed 15 March 2016). Ontario Energy Board, 2017. Natural gas rates – Historical. 〈http://www. ontarioenergyboard.ca/OEB/Consumers/Natural+Gas/Natural+Gas+Rates/Natural +Gas+Rates+-+Historical〉. (Accessed 12 July 2017). Ontario Ministry of Environment and Energy, 1996. A framework for competition. Report of the Advisory Committee on Competition in Ontario’s Electricity System to the Ontario Minister of Environment and Energy (May). Queen’s Printer for Ontario, Toronto. Rabin, M., Thaler, R.H., 2001. Anomalies: Risk aversion. J. Econ. Perspect. 15, 219–232. Rose, N.L., 2001. Regulation: Empirical analysis. In: Smelser, N.J., Baltes, P.B. (Eds.), first editionInternational Encyclopedia of the Social and Behavioral Sciences 19. Elsevier Science Ltd, Oxford, 12957–12963. Rose, N.L., 2014. Learning from the past: Insights for the regulation of economic activity. In: Rose, N.L. (Ed.), Economic Regulation and its Reform: What Have We Learned? NBER Conference Report. University of Chicago Press, Chicago, 1–23. Spence, D.B., 2008. Can law manage competitive energy markets? Cornell Law Rev. 93, 765–817. Stenson, F., 1985. Waste to wealth: a history of gas processing in Canada. Canadian Gas Processor Association/Canadian Gas Processors Suppliers’ Association, Calgary. Tirole, J., 2003. The Theory of Industrial Organization, Fourteenth Printing 1988. MIT Press, Cambridge. Tirole, J., 2015. Market failures and public policy. Am. Econ. Rev. 105, 1665–1682.

treating the distribution utility's gas commodity purchases as a cost to be passed on without markup already provided the consumer with the maximum surplus available in the commodity transaction. The commodity marketers’ own price data indicate that the lowest marketer price added roughly an additional three-quarters of the wholesale cost of gas to their customer's energy bill, compared to the utility's weighted average commodity cost. The data used to confirm the institutional prediction and to construct a conservative estimate of how much surplus marketers could divert from consumers was not available from official sources, nor indeed did marketers provide it as an historical time series. It was collected manually by checking the web postings over time. Another policy implication, then, is the importance of identifying, collecting and monitoring key data associated with a new set of institutional arrangements, in order for policy makers to make concrete statements about the benefits or lack thereof after a policy change. The price data collected here can of course now be applied to any confidential marketer sales data held by the regulator, to estimate the total loss to Ontario consumers due to the advent of natural gas marketing.23 Policy makers have focused for some time on the competitiveness of retail energy markets. As important as a market's competitiveness is to some questions, it is not relevant to others. The likely evolution of the size and distribution among market participants of surplus is perhaps the most important consideration in order to anticipate the implications of the policy change, or indeed to assess its ex post effectiveness. Although the entry of unregulated natural gas marketers might be thought to provide consumers with opportunities for lower gas commodity prices through retail competition, the marketers’ own pricing history provides no evidence to support this conjecture. Regulated utilities charge the actual average wholesale cost of natural gas purchased on the customer's behalf without marking it up, and earn their revenue requirements on carriage services. Natural gas marketers selling the commodity have no other activity from which to earn revenue to cover their costs. Under these institutional arrangements, regardless of whether they were pricing competitively or not, unregulated marketers could not have consistently offered a better commodity price to retail consumers than what has always been available from the utility. It is vital to consider the institutional arrangements carefully before arguing that competition is important or can improve the situation. Acknowledgements I thank Stuart Mestelman and David Feeny and two anonymous referees for helpful comments; and Umehani Ahmed for expert research assistance. Any remaining errors are my responsibility. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References Alberta Energy, 2017. Average Market Price (AMP), About Natural Gas: Prices. 〈http:// www.energy.alberta.ca/NaturalGas/1316.asp〉. (Accessed 20 March 2017). Arano, K.G., Velikova, M., 2009. Price convergence in natural gas markets: City-gate and residential prices. Energy J. 30, 129–154. Arano, K.G., Blair, B.F., 2008. An ex-post welfare analysis of natural gas regulation in the industrial sector. Energy Econ. 30, 789–806. Barcella, M.L., 1996. Natural gas in the twenty-first century. Bus. Econ. 31, 19–24. Barzel, Y., 1997. Economic Analysis of Property Rights second ed.. Cambridge University Press, Cambridge. Brau, R., Doronzo, R., Fiorio, C.V., Massimo, F., 2010. EU gas industry reforms and consumers' prices. Energy J. 31, 167–182. Canada. Department of Energy, Mines and Resources, 1985. Agreement (among the governments of Canada, Alberta, British Columbia, and Saskatchewan) on natural gas markets and prices. (Hallowe’en Agreement). Document # 85/162 (a) (including

23 I am indebted to an anonymous referee for pointing out that the Ontario Energy Board has collected data from marketers, and that a more refined consumer surplus loss

(footnote continued) estimate could be made using that confidential data.

564