Pipes, Trains and Automobiles: Explaining British Columbia's High Wholesale Gasoline Prices.

AuthorFellows, G. Kent

    In this paper I examine the effect of transportation constraints on imperfectly competitive regional wholesale fuel and crude oil markets. While the formal modelling exercise has potentially broad application, it is motivated here by the recent experience of western Canadian wholesale gasoline markets, where the demand for pipeline capacity from Alberta to British Columbia (BC) has exceeded new capacity investment over the last decade and a half. As suggested by the paper's title, this period of excess demand for pipeline capacity has been accompanied by a relative increase in the price of wholesale gasoline in BC markets. As such an empirical section following the formal model derivation tests and applies the model's analytical implications using panel data on western Canadian wholesale gasoline prices. The empirical results demonstrate the relationship between changes in the transportation supply curves and the observed wholesale prices within the context of a regional oligopoly market.

    While the empirical analysis is specific to western Canada, the analytical results can be extended to other regional markets for liquid or gaseous hydrocarbons as the analysis follows from a generalized extension of the well-known Cournot oligopoly model. Applications beyond hydrocarbon markets are possible as well. The model is general enough to apply to any oligopoly market wherein competitors face a common but discontinuous supply curve for a required input good. (1)

    The combination of a changing geographic pattern of North American production (the "U.S. shale-boom") and a lack of new investment in pipeline capacity has led to increased interest in inter-regional price differences and arbitrage conditions in crude and refined petroleum product markets (Agerton and Upton Jr, 2019; Borenstein and Kellogg, 2014; McRae, 2017; Oliver et al., 2014; Walls and Zheng, 2020). Existing work in the area has been primarily empirical, focusing on relationships between regional price series, with little or no attention paid to individual firm conduct. This level of abstraction is appropriate in investigating general regional pricing patterns, but insights on production and transportation cost pass through, refinery market power and competition policy issues require a more detailed examination of firm level conduct.

    This is particularly relevant in BC where wholesale price increases (relative to neighboring regions) and letters of comment alleging "collusion and price fixing" have recently motivated a formal BC Utilities Commission (BCUC) investigation into pricing and potential anti-competitive conduct among wholesale fuel suppliers (BCUC, 2019a,b). (2) As I argue here, these accusations are erroneous since these relative price increases can be explained by pipeline congestion rather than anti-competitive conduct when taking into account the oligopoly structure of the regional market. But the existence of a formal investigation by the BCUC does demonstrate the importance of understanding market power in regional markets facing potential transportation congestion.

    The role of supply capacity constraints in imperfectly competitive markets is well understood (Kreps and Scheinkman, 1983; Moreno and Ubeda, 2006). However insufficient pipeline capacity does not imply a fixed regional supply constraint at the firm or market level. Pipeline capacity is only one (generally the least expensive) mode of transportation for oil and gas commodities and refined products, but firms can employ higher cost alternatives. Given this, the appropriate model for regional crude or refined product market outcomes is a Cournot Oligopoly model in which all firms' face a common and potentially discontinuous supply function for inter-regional transportation, acknowledging lower cost (pipeline) and higher cost (rail, barge, trucking) substitutes. I develop such a model below and empirically test it using a natural experiment wherein a regulatory rule change dramatically reduced the available pipeline capacity for refined product shipments from the western Canadian refining hub in Alberta to markets in BC.

    The results of this exercise suggest that the pass-through rate of marginal transportation costs is not one-to-one, due to the imperfectly competitive nature of regional markets. It is also possible that reductions in pipeline capacity may reduce the rate of pass-through of other marginal cost components (i.e. those associated with production rather than transportation) if the supply curve for alternative transportation modes is upward sloping (i.e. not perfectly elastic). This combination of increased wholesale prices and reduced pass through of production costs can give the appearance of anti-competitive conduct (collusion and price fixing) even though firms' competitive strategies (response functions) remain unchanged. This is an important insight for regulators and competition policy authorities.

    I start with a brief overview of the BCUC Gasoline Price Investigation report before theoretically deriving and empirically testing the proposed Cournot model extension.


    Starting in mid-2015 BC gasoline prices began increasing relative to neighboring regions. This eventually prompted a BCUC investigation into wholesale gasoline prices and a two part report identifying an "unexplained difference of approximately 13 cpl [cents per liter]" between Vancouver and Seattle prices (BCUC, 2019a,b). (3)

    The BCUC's assessment implicitly uses a competitive arbitrage theory to determine a "correct" Vancouver price equal to the Seattle price plus transportation costs. Based on this counter-factual, they identify an average 13 cpl gap between the claimed "correct" price and the observed price, speculating that this gap indicates anti-competitive conduct. (4) Following the completion of the BCUC reports, the Government of BC more emphatically asserted that the price difference was the result of anti-competitive conduct and announced legislation intended to combat it (Government of British Columbia, 2019).

    However, as argued below, the relative price increase in BC markets can be explained by better acknowledging the role of inter-regional transportation costs (and specifically those related to pipeline capacity constraints) within the context of imperfectly competitive markets. The first step towards this is understanding how pipeline constraints have affected the available capacity for refined product transportation from the main western Canadian refining hub (in Edmonton Alberta) to BC regional markets.

    2.1 Pipeline Capacity Constraints and Apportionment

    As shown in Figure 1, the TransMountain Pipeline (TMPL) connects Edmonton to major regional BC markets (directly to Kamloops and Vancouver and indirectly to Nanaimo via shipments across the Strait of Georgia from Vancouver). It is the only pipeline that ships refined products to these BC markets and has been oversubscribed (excess demand relative to available capacity) since May 2010.

    Inter-provincial and international pipeline tolls and traffic are regulated in Canada by the National Energy Board (NEB).'' Under NEB regulation, when there is excess capacity demand, a pipeline's capacity is apportioned to individual shippers based on an apportionment rate [Please download the PDF to view the mathematical expression], where [n.sub.i], is the nominated volume (stated shipping demand) of an individual firm i and K is the existing pipeline capacity. Firms are then allocated individual capacity [k.sub.i] as per equation (1):

    [Please download the PDF to view the mathematical expression] (1)

    While the TMPL has been under semi-regular apportionment since at least 2006, the apportionment rate on the pipeline grew significantly in 2010 and was persistently above 0.6 (60%) from mid-2011 until 2015 (Figure 2). (6) It is worth noting here that the excess demand for capacity on the TMPL has occurred despite a clear effort to increase pipeline capacity between Edmonton and the Vancouver area. The proposed Trans-Mountain Expansion (TMX) will, once completed, more than double the throughput capacity on the system. However, TMX has faced significant delays caused by a confluence of "... legal risk and social mobilization and a major regulatory barrier" (Janzwood, 2020).

    In 2012 Chevron Canada filed a complaint with the NEB alleging that TMPL's nomination verification procedure allowed for "strategic over-nomination" (7) by some shippers creating an unfair preference in capacity allocations. The NEB investigation following from this complaint determined that TMPL's procedures were "likely contributing to the ongoing apportionment of the Pipeline" because shippers may have been able to inflate nominations to secure larger proportions of pipeline capacity (NEB. 2013). Given the apportionment allocation formula it is clear that if a shipper expects over-subscription, they have an incentive to strategically over-nominate in order to avoid the effects of apportionment and obtain a larger capacity allocation. TMPL historically attempted to guard against strategic over-nomination by requiring third-party verification that a shipper was capable and prepared to fully ship the entire nominated volume.

    As a result of its 2012 investigation the National Energy Board directed TMPL to modify its nomination verification procedure in order to reduce apportionment. This was an implicit order to reduce strategic over-nomination since the underlying demand and the supply of capacity should be unaffected by nomination rules. (8) The new verification procedure, brought into force in May 2015, limits nominated volumes to be the higher of: "...average deliveries to a shipper... using [the highest] 18 months out of the 24-month period immediately preceding the nomination date; or three per cent of the available capacity" (NEB. 2015). This led to a dramatic reduction in...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT