Determinants of fuel prices: dominant firms, local monopolies and "captive" demand

Determinants of fuel prices: dominant firms, local monopolies and ‘captive’demand Luis Ángel Hierro-Recio a , Pedro Atienza-Montero b , María Varo- Morales c and Antonio José Garzón-Gordón d ABSTRACT This paper analyses the effect on retail fuel prices of factors such as belonging to dominant firms, the position of a local monopoly or oligopoly, and service station location. We study the effect of belonging to the dominant firms in the market, Repsol and Cepsa, of enjoying a local natural monopoly or oligopoly in rural areas, and of being located in places with captive demand, such as highways and motorways, as well as in the city centre. We apply this study to service stations in the province of Seville (Spain). The main findings are that the two main distributors, Repsol and Cepsa, set a higher price. We also find market power at a local level, which appears through monopoly or duopolies in rural areas, and which also results in higher prices, albeit to a much lesser degree. In addition, we see that stations servicing users on high-capacity roads as well as stations located in Seville city centre also set higher prices. ARTICLE HISTORY Received 13 September 2019; Accepted 30 July 2020 KEYWORDS service stations; fuel price; market dominance; captive demand JEL CLASSIFICATIONS D42; D43; L13 INTRODUCTION Throughout its recent history, the retail fuel distribution market in Spain has seen major state intervention and high concentration. 1 Between 1927 and 1984, the state held a monopoly in the sector, managed by the Compañía Arrendataria del Monopolio de Petróleos Sociedad Anónima (CAMPSA), which imported, refined, stored, distributed and sold petroleum products, in addition to engaging in exploration and producing hydrocarbons. In 1984, before joining the European Economic Community (EEC), Spain initiated a transitional process of demonopolization, regulated under Law 45/84, governing the reorganization of the oil sector, which lasted until 1992, during which the system of fixed prices was replaced by a system of maximum prices. © 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/ by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. CONTACT [email protected] aDepartment of Economics and Economic History, Faculty of Economics and Business, University of Seville, Seville, Spain. (Corresponding author) [email protected] bDepartment of Economics and Economic History, Faculty of Economics and Business, University of Seville, Seville, Spain. [email protected] cDepartment of Economics and Economic History, Faculty of Economics and Business, University of Seville, Seville, Spain. [email protected] dDepartment of Economics and Economic History, Faculty of Economics and Business, University of Seville, Seville, Spain. REGIONAL STUDIES, REGIONAL SCIENCE 2020, VOL. 7, NO. 1, 394–411 https://doi.org/10.1080/21681376.2020.1811138 In 1992, Law 34/1992, governing the Organization of the Oil Sector dissolved CAMPSA’s state monopoly, segregating the company’s activities and liberalizing all the activities which until then had been managed by public monopoly, except management of the transport and storage network, which came under the supervision of the newly created Compañía Logística de Hidrocarburos (CLH). Full liberalization of fuel prices came about in 1996 and for other fuels in 1998. Despite the sector’s intense transformation over the last two decades, which has witnessed liberalization and privatization, there are clear indications of significant market power. In terms of production, of the nine refineries in Spain, eight are owned by the two largest Spanish oil companies (Repsol owns five and Cepsa owns three), with the other being owned by British Petroleum (BP). In terms of distribution, of the 7631 stations in Spain in 2015, 3544 (46.4%) were Repsol and 1512 (19.8%) were Cepsa. In other words, Repsol and Cepsa monopolized two-thirds (66.6%) of all service stations in the country in 2015, and if we add BP (8.2% of stations), Galp (7.5%) and Disa (7.2%), the top five brands cornered 90% of all service stations (Spanish Association of Operators of Petroleum Products, 2015). Such a high level of market concentration substantially increases the likelihood of collusive duo- or oligopoly that would lead to higher prices (Borenstein & Shepard, 1996; Castanias & Johnson, 1993; Haltiwanger & Harrington, 1991; Maskin & Tirole, 1988; Rotemberg & Saloner, 1986; Slade, 1987,1992). This behaviour can be reproduced in the case of small areas. In rural areas, reduced demand can lead to natural local monopoly or duopolies, and fuel prices may be higher. Together with the market power afforded by a concentration of supply in the hands of just a few companies, the location of stations plays a key role in the retail fuel market, and can also lead to higher prices. Models of spatial competition, such as Haining (1983,1985), Haining et al. (1996), Anderson and de Palma (1992), Capozza and Van Order (1989), Greenhut et al. (1987) or Norman (1986), have been taken as a basis to study this issue. With regard to higher prices linked to the location of stations, some researchers posit the analysis as a problem of product differentiation 2 and, therefore, as a problem of monopolistic market competition, explored in models such as those of Hotelling (1929), Salop (1979), Perloff and Salop (1985) or Anderson et al. (1992), among others. Our objective in this work is to jointly present different factors that may lead to higher fuel prices associated with firms’market dominance and location factors, in an attempt to determine how much more consumers pay due to these factors associated with market structure. In particular, we estimate the effect of belonging to a firm with a high market share, Repsol and Cepsa, of being a local monopoly or oligopoly in a rural area, and of being located on high-capacity roads, that is, highways and motorways, and in Seville city centre, where demand is captive. As control variables, we include the fleet of vehicles registered in the municipality where the station is located, as a proxy of demand factors; the number of stations within a radius of 2 km, controlling for spatial competition in the market and, as a robustness test, we also include crude oil prices. For our work, we use the net tax price of gasoline 95 and diesel at service stations in the province of Seville from 26 March to 21 July 2015. We excluded stations as well as days for which we did not have full daily price data for both fuels. This led us to remove 13 service stations out of a total of 353 stations involved in the study period, and 24 days of the period considered, leaving a 94-day sample. We estimate a panel data model with time-specificfixed effects in order to control for the variation in wholesale costs and temporary fluctuations in demand. In addition, as a robustness test, we perform two additional estimates using panel-corrected standard errors (PCSE) estimated by Prais– Winsten regression, and feasible generalized least squares (FGLS), including delayed oil prices, which, in addition to confirming the previous results, provides ahighdegreeofexplanatorypower. The dependent variables are the daily prices of gasoline 95 and diesel A, with the variables representing potential sources of higher prices being incorporated in the form of dummy variables. Determinants of fuel prices: dominant firms, local monopolies and ‘captive’demand 395 REGIONAL STUDIES, REGIONAL SCIENCE The previous literature exploring price differences in stations in Spain is scarce and includes only four studies: Bello Pintado and Cavero Brújula (2007,2008), Bello Pintado and Contín- Pilart (2010) and Albalate and Perdiguero (2015). The first three examine the impact of brands with refineries in Spain, namely Repsol, Cepsa and BP, density of stations by regions, those within 2 km, or the distance to the nearest station. As regards location, they analyse whether the station is located on a toll road, a national road, in the city centre and peripheral areas of the city with large distribution centres. However, Albalate and Perdiguero (2015) is much more specific, since it focuses on the difference in fuel prices of stations located on toll motorways compared with those located on toll-free motorways. The aim of our work is to add fresh empirical literature for Spain in the field. The main results we obtain from our analysis are: the two main distributors, Repsol and Cepsa, with dominant positions in the market, set higher prices; and that local monopoly or duopolies in rural areas also set higher prices, although these are smaller. As for the effect of location, prices are set higher in stations serving users of high-capacity roads, such as toll motorways and free motorways. Service stations located within the ring formed by the SE-30 ring road and the northern ring road in the city of Seville also set higher prices. The paper is organized as follows. The next section reviews the previous literature. The third section presents the methodology and describes the variables used, as well as the data processing. The fourth section presents the results. Finally, in the fifth section concludes. REVIEW OF THE EMPIRICAL LITERATURE The empirical literature on retail fuel prices has experienced significant growth over the last 20 years. Following Eckert (2013), said literature can be classified into two main groups: empirical studies of fuel price dynamics; and studies of the determinants of retail prices and price differences. The first group contains studies examining the asymmetric response of retail prices to wholesale prices or the price of crude oil and the literature of the Edgeworth price cycles. As regards the former, it has empirically been shown that the response is faster to price increases than to decreases (‘rocket and feather effect’). The classic work on this subject is Borenstein et al. (1997). As for the Edgeworth cycles, which deal with rises and falls in prices and which occur cyclically, the phenomenon is formalized theoretically by Maskin and Tirole (1988), although these have their origin in Edgeworth (1925), and have led to extensive empirical research (Atkinson, 2009; Castanias & Johnson, 1993; Eckert, 2002,2003; Lewis, 2012; Noel, 2007; Wang, 2008; Zimmerman et al., 2010; or more recent works by Byrne et al., 2015; Foros & Steen, 2013; and Isakower & Wang, 2014, among others). Within the second group of works, related to determinants of retail prices and price differences and within which the present work is framed, Eckert (2013) distinguishes four types of empirical studies: the first involves estimates of reduced-form price equations which relate prices to market structure at the national or municipal level (Chouinard & Perloff, 2007; Sen, 2003,2005); the second involves empirical analyses of the effects of mergers and regulations on prices, normally using difference-in-difference estimates (Coloma, 2002; Hastings, 2004; Silvia & Taylor, 2010; Taylor et al., 2010, among others); the third deals with structural estimates of market power (Houde, 2012; Manuszak, 2010); and the fourth addresses estimates of price determinants at the station level, as well as their dispersion and uniformity. Our work is framed within this latter group of studies that seek to analyse price determinants. For the purposes of this work, these factors can be grouped according to market concentration, the degree of spatial competence or the spatial differentiation of the product. 396 Luis Ángel Hierro-Recio et al. REGIONAL STUDIES, REGIONAL SCIENCE In terms of market concentration, studies tend to consider the existence of one or more main brands. The conclusion to emerge from most works is that stations belonging to brands with the highest number of service stations set higher fuel prices. This is reported in Van Meerbeeck (2003) for Shell throughout Belgium; in Barron et al. (2004) for Chevron, Exxon Mobil/BP, Shell, Texaco and Unocal in the United States (San Francisco urban areas, Tucson, Phoenix and San Diego); in Haucap et al. (2017) for Aral, Shell, Esso, Total and Jet in Germany; in Cooper and Jones (2007) for BP, Chevron, Exxon, Marathon or Shell in Lexington (Kentucky); in Bello Pintado and Contín-Pilart (2010) for BP and Gulf in Spain; and in Bello Pintado and Cavero Brújula (2007,2008) for Repsol, Cepsa and BP for 22 Spanish provinces. In contrast, the work of Remer (2019) in Kentucky and Virginia concludes that the margins are lower in the stations of major brands. Together with the above, another type of work analyses the effects on fuel prices of indicators that are representative of the degree of market concentration. Clemenz and Gugler (2009) use the Herfindahl index 3 and the concentration percentage of the main brand and the four major oil companies in 2856 stations in Austria, and conclude that concentration increases price. The same conclusion is reached by Haucap et al. (2017) for a sample of stations in Germany, and by Remer (2019) for Kentucky and Virginia, using market participation indicators. Kihm et al. (2016) use the Herfindahl index in interaction with the Brent crude price for a sample of 13,000 stations in Germany, and again conclude that market concentration bumps up prices. Certain works use opposite indicators to market concentration, such as opening out of the market, by counting the presence of independent companies, such as Clemenz and Gugler (2009), with their conclusion being that the participation of independent companies drives prices down, and Haucap et al. (2017), who obtain the opposite result. The analysis of factors related to spatial competition encompasses a wide range of works. In order to identify the effect of competition, different types of variables are used. One group of papers uses variables representative of the density of stations over a given area. Most of these studies report that the higher the density of stations, the lower the prices, which would indicate that competition pushes prices down. A range of variables is used: Van Meerbeeck (2003) uses the number of competing stations in the same municipality; Barron et al. (2004) use the number of stations within 1.5 miles (2.4 km); Clemenz and Gugler (2009) use the number of stations per square kilometre; Haucap et al. (2017) use the number of competitors within 2 km; Remer (2019) use the number of stations in each county of Virginia and Kentucky; Bello Pintado and Cavero Brújula (2007) use the number of stations in the same region; and Cooper and Jones (2007) use the number of stations to the central business district of Lexington (Kentucky) and the number of stations located outside commuter roads. However, there is a lack of consensus in the results. In Remer (2019), who uses the number of competing stations within a 1.5-mile variable, and Bello Pintado and Contín-Pilart (2010), who use the number of competitors within a 2 km variable, density is seen to increase prices, whereas in Hosken et al. (2008), using the number of stations within a radius of 1.5 miles has no statistically significant effect. A second type of variable used in the literature is distance to the nearest stations. In these works, the majority result is that the greater the distance, the higher the price. Such is the case in Barron et al. (2004), Haucap et al. (2017), Cooper and Jones (2007) and Bello Pintado and Cavero Brújula (2008). However, the result in Hosken et al. (2008) is not statistically significant. One variation of this type of variable is to measure the distance in travel time or travel cost to the nearest station. When using this variation, neither Clemenz and Gugler (2009) nor Hosken et al. (2008)find the variables to be statistically significant. Finally, when analysing spatial location and using it as a differentiating element, the conclusion to emerge is that price is higher: if the station is located on a toll road (Bello Pintado & Cavero Brújula, 2007,2008; Haucap et al., 2017); the station is located on a national road Determinants of fuel prices: dominant firms, local monopolies and ‘captive’demand 397 REGIONAL STUDIES, REGIONAL SCIENCE or in a city centre (Bello Pintado & Cavero Brújula, 2007,2008); and if the station is located on a ring road, on a non-main road, at a supermarket or on the corner between two roads (Ning & Haining, 2003). 4 METHODOLOGY AND DATA Our objective is to verify empirically the existence of higher prices in the retail fuel sector deriving from belonging to a firm that dominates the market, resulting from either the monopolistic or oligopolistic structure of the market, both in the whole of the geographical area considered and at the level of small localities, or through being located where demand is captive. For this, we take a sample of prices from stations in the province of Seville for the period 26 March–21 July 2015, and which excludes 24 days 5 due to lack of available data for those dates, such that the study covers a total of 94 days. During that period in the province of Seville there were 353 stations, 6 of which we removed 13 stations 7 that did not simultaneously supply the price of diesel and gasoline or for which not all the daily fuel prices were available and it was not possible to perform interpolation. For the final 340 stations, see Table A1 in Appendix A in the supplemental data online. For stations with incomplete data, and whenever possible, we performed interpolation between the previous and subsequent price in order to determine the price on days that lacked data. This was the most common data-processing procedure used and was applied to 37 stations. 8 For stations where prices varied very infrequently, we assumed that prices did not change until the next recorded price difference, and we used the price before the next data variation. This was applied to prices at nine stations. 9 Given the high level of taxation affecting the sector, we felt it preferable to use the net tax price. Fuel taxation in 2015 consisted of the state tax on hydrocarbons (general and special type), regional levy of the same tax, and value added tax (VAT), which is applied to the net tax price on fuels plus the previously mentioned hydrocarbon tax rates. We first deducted VAT (the standard 21%) and subsequently deducted the hydrocarbon tax. In our case, for gasoline 95, it is a general state tax rate of €400.49 per 1000 litres and a special tax rate of €24 per 1000 litres, to which must be added €48 regional levy per 1000 litres. As for diesel A, in 2015, the general rate of hydrocarbon tax was €307 per 1000 litres, the special rate was €24 and the regional tax was €48 per litre. 10 The econometric specifications of the model for the two dependent variables of gasoline and diesel in each scenario are: Scenario 1: dieselprice i,t (gasolineprice i,t )=β 0 +δ t +β 1 repsol-cepsa i,t +β 2 localmonopoly/duopoly i,t +β 3 motorways i,t +β 4- citycentre i,t +β 5 stationdensity i,t +β 6 vehicles1(2) i,t +u i,t Scenario 2: dieselprice i,t (gasolineprice i,t )=β 0 +δ t +β 11 repsol i,t +β 12 cepsa i,t +β 2 localmonopoly/duopoly i,t +β 3 motorways i,t + β 4 citycentre i,t +β 5 stationdensity i,t +β 6 vehicles1(2) i,t +u i,t where repsol-cepsa i,t is a dummy variable that takes the value 1 if the station is owned by Repsol or Cepsa, and 0 if owned by a different brand; repsol i,t is a dummy variable that takes the value 1 if the station is owned by Repsol, and 0 if owned by a different brand; cepsa i,t is a dummy variable that takes the value 1 if the station is owned by Cepsa, and 0 if owned by a different brand; localmonopoly/duopoly i,t refers to stations with a local monopolistic or oligopolistic position –it is a dummy variable that takes the value 1 if it is the only station in the area or, at most, competes with one other station in the same area, and takes 0 for the rest; motorways i,t is a dummy variable 398 Luis Ángel Hierro-Recio et al. REGIONAL STUDIES, REGIONAL SCIENCE that takes the value 1 when the station is on toll motorways or toll-free motorways, and 0 when located on other types of road; citycentre i,t is a dummy variable equal to 1 if the station is in the centre of Seville, which is deemed to be the area within the confines of the city ring road, the SE- 30; stationdensity i,t is defined as the number of stations within a radius of 2 km, as a proxy for spatial competition; vehicles 1(2) i,t is the number of registered vehicles in the municipality where the station is located; diesel or gasoline engine, respectively; β 0 is the constant term; and δ t is time-specificfixed effects. By using the variable repsol-cepsa, our aim is to capture the existence of higher prices derived from the market structure and the market power these brands might have when compared with other stations. The percentage of stations belonging to these two brands, which already represents a majority in Spain (66.6%), reached 75.6% in the province of Seville (42.4% the former and 33.2% the latter). This is followed, quite some way behind, by BP (10.4%) and Galp (7.3%) (Spanish Association of Operators of Petroleum Products, 2015). As anticipated by the theoretical and empirical literature, the stations of these main brands establish a significantly higher fuel price. Additionally, we estimate the effects of Repsol and Cepsa separately in order to capture the possibility that Repsol is acting as leader of the duopoly. With the dummy variable localmonopoly/duopoly our aim is to capture the effect on the price of being in a monopolistic or oligopolistic position in small towns, when the stations are the sole suppliers or, at most, when they share the local market with one other station. This type of station is located in areas with little traffic, such as in small towns found mainly in the country, or in the mountains to the north and south of the province. There are 66 such towns and 88 stations involved (Figure 1). 11 This variable has not been used in the previous literature. However, it is expected to have a similar effect to that of the previous variable, such that in these stations the price should be higher. The closest variables employed in the literature are the number of local competitors, used by Van Meerbeeck (2003), and the number of stations in a county, used by Remer (2019). In both studies, the relationship with price proves to be negative; the greater the number of competitors, the lower the price, which is consistent with the expected effect described for our variable. With regard to the variables designed to capture higher prices associated to the location of service stations, the variable motorways aims to detect the possible existence of higher prices associated with long-distance travel, where not knowing the locations of the stations and the price differences between them makes drivers tend to refuel at the station adjacent to the toll and toll-free motorways. This is what we might refer to as ‘trapped’demand (Albalate & Perdiguero, 2015). As we have seen, the empirical literature has included this variable and has confirmed its positive impact on fuel price (Bello Pintado & Cavero Brújula, 2007,2008; Bello Pintado & Contín-Pilart, 2010; Haucap et al., 2017). 12 In this paper, we therefore expect a similar effect. In Seville, there are 61 service stations located in the vicinity of this type of road (Figure 2). 13 With the dummy variable citycentre, we assign a value of 1 to stations located within the area bounded by the Seville ring road, the SE-30, and the Ronda Urbana Norte (Figure 3). 14 With this variable, we aim to capture the effect of a type of demand, which, in a certain sense, is also ‘trapped’by the difficulty of travelling within the city, by the massive influx of vehicles and by the small number of stations compared with the large number of trips made. This variable has been used by Bello Pintado and Cavero Brújula (2007) and we expect to obtain a similar result of higher fuel prices. As control variables, we first include vehicles1 and vehicles2, representing the number of registered vehicles in the municipality where the station is located (we distinguish between diesel (1) and gasoline (2) engine cars depending on the type of fuel prices we are estimating), as a proxy of demand factors. 15 Finally, we include a factor representative of the degree of spatial competition: the density of stations within a radius of 2 km, which is a variable commonly used in empirical Determinants of fuel prices: dominant firms, local monopolies and ‘captive’demand 399 REGIONAL STUDIES, REGIONAL SCIENCE literature (Barron et al., 2004; Bello Pintado & Contín-Pilart, 2010; Haucap et al., 2017; Hosken et al., 2008; Ning & Haining, 2003; Remer, 2019). 16 Additionally, as a robustness test, we include the Brent oil price, which is the cost of basic fuel production and the benchmark for European markets. We use the variable with a one- and twoday delay since, as Borenstein and Shepard (1996) point out, companies do not react to changes in oil price immediately. A delay of at least one day is required to adjust fuel prices and, bearing in mind that intermediary refining companies are also involved, at least one further day’s delay is needed (Borenstein & Shepard, 1996). Model variables, including those in the robustness test, and their sources are shown in Table 1. To allocate the values of the dummy variables on location, we use the coordinates of each station provided by the Geoportal web. RESULTS For our analysis, we use a panel database that combines the temporal dimension (fuel price for 94 days) and the cross-section (340 stations in the province of Seville). This is a balanced panel, such that each station has the same number of observations for all variables. The econometric method applied consists of a panel data model with time-specific fixed effects to control for wholesale cost variations, mainly due to oil price variations, as is Figure 1. Station localization: local monopoly/oligopoly. 400 Luis Ángel Hierro-Recio et al. REGIONAL STUDIES, REGIONAL SCIENCE common in the literature. Furthermore, we also estimate a different specification as a robustness test. Table 2 shows the descriptive statistics for each variable. The fuel price range, namely the difference between the maximum and minimum, reveals there was considerable price variation in the 94 days analysed. For gasoline, the percentage difference is 48.8% between the minimum and maximum, while for diesel it is 47.3%. The Brent oil price difference between the minimum and maximum was 23.54%. That is, the time sample corresponds to a period which, due to price oscillation, allows us to infer significant price-variation behaviour. Table 3 shows the results obtained when applying panel data with time-specificfixed effects. The estimates of the two specifications, diesel and gasoline, and in the two scenarios, are jointly statistically significant, as shown by the Chi 2 Wald test with a p-value of 0. Likewise, R 2 is around 0.40 in the first scenario and 0.31 in the second scenario. In order to enhance the robustness of the results, Table 4 shows the results obtained from the PCSE and FGLS estimation, as explained above. All the estimates in the robustness test are jointly significant, with a Wald Chi 2 test pvalue of 0. It should also be noted that the R 2 in these estimates is much higher than in the baseline estimates using time fixed effects, and now reaches an R 2 of 0.81 in the case of diesel and 0.75 in the case of gasoline. Finally, it is worth noting that the estimated coefficients of the variables are practically the same when using both time fixed effects as well as PCSE and FGLS. Almost all the variables considered as determinants of higher prices are statistically significant at 1%. 17 Figure 2. Station localization: toll and toll-free motorways. Determinants of fuel prices: dominant firms, local monopolies and ‘captive’demand 401 REGIONAL STUDIES, REGIONAL SCIENCE If we consider the variable ‘Repsol-Cepsa’, the results show that they charge a price of 2.8 (diesel A) and 2.7 (gasoline 95) cents higher than the average of all the service stations. When separating the two brands ‘Repsol’and ‘Cepsa’, we found that Repsol sets the highest price which is, on average, 2.1 and 2.0 cents higher, respectively, whereas Cepsa stations set higher but significantly lower prices than Repsol. Cepsa sets a price per litre that is 0.6 cents higher for diesel and 0.3 cents higher for gasoline. Service stations enjoying a local monopolistic or oligopolistic position set a price that is between 0.25 and 0.35 cents higher than other service stations and between 0.18 and 0.25 cents higher in the case of diesel. With regard to location, the highest prices are those obtained by stations located in Seville city centre, whereas service stations located on toll or toll-free motorways establish a price that is 0.9 cents higher for diesel as well as for gasoline. As regards the control variables, the fleet of registered vehicles according to the type of fuel, which represents demand factors, has a coefficient that is found to be positive and statistically significant. With regard to spatial competition, which is represented by the number of stations Figure 3. Station localization: city centre. 402 Luis Ángel Hierro-Recio et al. REGIONAL STUDIES, REGIONAL SCIENCE Atkinson, B. (2009). Retail gasoline price cycles: Evidence from Guelph, Ontario using bi-hourly, station-specific retail price data. The Energy Journal,30(1), 85–109. https://doi.org/10.5547/ISSN0195-6574-EJ-Vol30- No1-4 Barron, J. M., Taylor, B. A., & Umbeck, J. R. (2004). Number of sellers, average prices and price dispersion. International Journal of Industrial Organization,22(8–9), 1041–1066. https://doi.org/10.1016/j.ijindorg. 2004.05.001 Bello Pintado, A., & Cavero Brújula, S. (2007). Competencia estratégica en la distribución minorista de combustibles de automoción. Revista de economía aplicada,15(45), 125–154. Bello Pintado, A., & Cavero Brújula, S. (2008). The Spanish retail petroleum market: New patterns of competition since the liberalization of the industry. Energy Policy,36(2), 612–626. https://doi.org/10.1016/j. enpol.2007.10.014 Bello Pintado, A., & Contín-Pilart, I. (2010). Influencia de los factores de localización en la fijación de los precios de los carburantes de automoción en España. Cuadernos Económicos de ICE,79,45–67. Borenstein, S., Cameron, A. C., & Gilbert, R. (1997). Do gasoline prices respond asymmetrically to crude oil price changes? The Quarterly Journal of Economics,112(1), 305–339. https://doi.org/10.1162/ 003355397555118 Borenstein, S., & Shepard, A. (1996). Dynamic pricing in retail gasoline markets. The RAND Journal of Economics, 27(3), 429–451. https://doi.org/10.2307/2555838 Byrne, D. P., Leslie, G. W., & Ware, R. (2015). How do consumers respond to gasoline price cycles? The Energy Journal, 115–147. https://doi.org/10.5547/01956574.36.1.5 Capozza, D. R., & Van Order, R. (1989). Pricing under spatial competition with consistent conjectures. Journal of Regional Science,29(1), 1–13. https://doi.org/10.1111/j.1467-9787.1989.tb01218.x Castanias, R., & Johnson, H. (1993). Gas wars: Retail gasoline price fluctuations. The Review of Economics and Statistics,75(1), 171–174. https://doi.org/10.2307/2109643 Chouinard, H. H., & Perloff, J. M. (2007). Gasoline price differences: Taxes, pollution regulations, mergers, market power, and market conditions. The B.E. Journal of Economic Analysis & Policy,7(1). https://doi.org/10. 2202/1935-1682.1599 Clemenz, G., & Gugler, K. (2009). Locational choice and price competition: Some empirical results for the Austrian retail gasoline market. In G. Arbia & B. H. Baltagi (Eds.), Spatial Econometrics (pp. 223–244). Physica-Verlag HD. Coloma, G. (2002). The effect of the Repsol–YPF merger on the Argentine gasoline market. Review of Industrial Organization,21(4), 399–418. https://doi.org/10.1023/A:1021122815423 Cooper, T. E., & Jones, J. T. (2007). Asymmetric competition on commuter routes: The case of gasoline pricing. Southern Economic Journal,74, 483–504. https://doi.org/10.2307/20111978 Eckert, A. (2002). Retail price cycles and response asymmetry. Canadian Journal of Economics/Revue Canadienne d’Economique,35(1), 52–77. https://doi.org/10.1111/1540-5982.00120 Eckert, A. (2003). Retail price cycles and the presence of small firms. International Journal of Industrial Organization,21(2), 151–170. https://doi.org/10.1016/S0167-7187(02)00038-3 Eckert, A. (2013). Empirical studies of gasoline retailing: A guide to the literature. Journal of Economic Surveys,27 (1), 140–166. https://doi.org/10.1111/j.1467-6419.2011.00698.x Edgeworth, F. Y. (1925). Papers relating to political economy (Vol. 1). MacMillan. Foros, Ø, & Steen, F. (2013). Vertical control and price cycles in gasoline retailing. The Scandinavian Journal of Economics,115(3), 640–661. https://doi.org/10.1111/sjoe.12024 Greenhut, M. L., Norman, G., & Hung, C.-S. (1987). The economics of imperfect competition. Cambridge University Press. Haining, R. (1983). Modeling intraurban price competition: An example of gasoline pricing. Journal of Regional Science,23(4), 517–528. https://doi.org/10.1111/j.1467-9787.1983.tb01007.x Haining, R. (1985). The spatial structure of competition and equilibrium price dispersion. Geographical Analysis, 17(3), 231–242. https://doi.org/10.1111/j.1538-4632.1985.tb00843.x Determinants of fuel prices: dominant firms, local monopolies and ‘captive’demand 409 REGIONAL STUDIES, REGIONAL SCIENCE Haining, R., Plummer, P., & Sheppard, E. S. (1996). Spatial price equilibrium in interdependent markets: Price and sales configurations. Papers in Regional Science,75(1), 41–64. https://doi.org/10.1007/BF02406959 Haltiwanger, J., & Harrington, J. E. Jr. (1991). The impact of cyclical demand movements on collusive behavior. The RAND Journal of Economics,22(1), 89–106. https://doi.org/10.2307/2601009 Hastings, J. S. (2004). Vertical relationships and competition in retail gasoline markets: Empirical evidence from contract changes in southern California. American Economic Review,94(1), 317–328. https://doi.org/10.1257/ 000282804322970823 Haucap, J., Heimeshoff, U., & Siekmann, M. (2017). Fuel prices and station heterogeneity on retail gasoline markets. The Energy Journal,38(6), 81–103. https://doi.org/10.5547/01956574.38.6.jhau Hosken, D. S., McMillan, R. S., & Taylor, C. T. (2008). Retail gasoline pricing: What do we know? International Journal of Industrial Organization,26(6), 1425–1436. https://doi.org/10.1016/j.ijindorg.2008.02.003 Hotelling, H. (1929). Stability in competition. The Economic Journal,39(153), 41–57. https://doi.org/10.2307/ 2224214 Houde, J. F. (2012). Spatial differentiation and vertical mergers in retail markets for gasoline. American Economic Review,102(5), 2147–2182. https://doi.org/10.1257/aer.102.5.2147 Isakower, S., & Wang, Z. (2014). A comparison of regular price cycles in gasoline and liquefied petroleum gas. Energy Economics,45, 445–454. https://doi.org/10.1016/j.eneco.2014.08.014 Kihm, A., Ritter, N., & Vance, C. (2016). Is the German retail gasoline market competitive? A spatial–temporal analysis using quantile regression. Land Economics,92(4), 718–736. https://doi.org/10.3368/le.92.4.718 Lewis, M. S. (2012). Price leadership and coordination in retail gasoline markets with price cycles. International Journal of Industrial Organization,30(4), 342–351. https://doi.org/10.1016/j.ijindorg.2011.12.002 Manuszak, M. D. (2010). Predicting the impact of upstream mergers on downstream markets with an application to the retail gasoline industry. International Journal of Industrial Organization,28(1), 99–111. https://doi.org/ 10.1016/j.ijindorg.2009.07.002 Maskin, E., & Tirole, J. (1988). A theory of dynamic oligopoly, II: Price competition, kinked demand curves, and Edgeworth cycles. Econometrica,56(3), 571–599. https://doi.org/10.2307/1911701 National Energy Commission. (2006). Cronología del Sector Petrolero español. Retrieved March 13, 2019 from https://petroleoyretail.com/wp-content/uploads/2013/04/crono-petroleo-CNE.pdf.pdf Ning, X., & Haining, R. (2003). Spatial pricing in interdependent markets: A case study of petrol retailing in Sheffield. Environment and Planning A: Economy and Space,35(12), 2131–2159. https://doi.org/10.1068/a3636 Noel, M. D. (2007). Edgeworth price cycles, cost-based pricing, and sticky pricing in retail gasoline markets. Review of Economics and Statistics,89(2), 324–334. https://doi.org/10.1162/rest.89.2.324 Norman, G. (Ed.). (1986). London papers in regional science 16. Spatial pricing and differentiated markets. Pion. Perloff, J. M., & Salop, S. C. (1985). Equilibrium with product differentiation. The Review of Economic Studies,52 (1), 107–120. https://doi.org/10.2307/2297473 Pesaran, M. H. (2004). General diagnostic tests for cross section dependence in panels. Cambridge Working Papers in Economics No. 0435, Faculty of Economics, University of Cambridge. Remer, M. (2019). Competition and the complexity of pricing strategies: evidence from retail gasoline stations. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2819138 Rotemberg, J., & Saloner, G. (1986). A supergame-theoretic model of price wars during booms. American Economic Review,76, 390–407. Salop, S. C. (1979). Monopolistic competition with outside goods. The Bell Journal of Economics,10(1), 141–156. https://doi.org/10.2307/3003323 Sen, A. (2003). Higherprices at Canadian gas pumps: International crude oil prices orlocal marketconcentration?An empirical investigation. Energy Economics,25(3), 269–288. https://doi.org/10.1016/S0140-9883(02)00097-X Sen, A. (2005). Does increasing the market share of smaller firms result in lower prices? Empirical evidence from the Canadian retail gasoline industry. Review of Industrial Organization,26(4), 371–389. https://doi.org/10. 1007/s11151-005-4206-4 Shepard, A. (1993). Contractual form, retail price, and asset characteristics in gasoline retailing. The RAND Journal of Economics,24(1), 58–77. https://doi.org/10.2307/2555953 410 Luis Ángel Hierro-Recio et al. REGIONAL STUDIES, REGIONAL SCIENCE Silvia, L., & Taylor, C. T. (2010). Petroleum mergers and competition in the northeast United States. Federal Trade Commission Working Paper 300. Slade, M. E. (1987). Interfirm rivalry in a repeated game: An empirical test of tacit collusion. The Journal of Industrial Economics,35(4), 499–516. https://doi.org/10.2307/2098585 Slade, M. E. (1992). Vancouver’s gasoline-price wars: An empirical Exercise in Uncovering Supergame Strategies. The Review of Economic Studies,59(2), 257–276. https://doi.org/10.2307/2297954 Spanish Association of Operators of Petroleum Products. (2015). Memoria, 2015. Retrieved March 3, 2019 from http://www.aop.es/media/1575/aop-alta.pdf Taylor, C. T., Kreisle, M. M., & Zimmerman, P. R. (2010). Vertical relationships and competition in retail gasoline markets: Empirical evidence from contract changes in southern California: Comment. American Economic Review,100(3), 1269–1276. https://doi.org/10.1257/aer.100.3.1269 Tirole, J. (1988). A theory of industrial organization. MIT Press. Van Meerbeeck, W. (2003). Competition and local market conditions on the Belgian retail gasoline market. De Economist,151(4), 369–388. https://doi.org/10.1023/B:ECOT.0000006590.66223.9a Wang, Z. (2008). Collusive communication and pricing coordination in a retail gasoline market. Review of Industrial Organization,32(1), 35–52. https://doi.org/10.1007/s11151-008-9163-2 Wooldridge, J. M. (2002). Econometric analysis of cross section and panel data. MIT Press. Zimmerman, P. R., Yun, J. M., & Taylor, C. T. (2010). Edgeworth price cycles in gasoline: Evidence from the U.S. Federal Trade Commission Bureau of Economics, Working Paper No 303. Determinants of fuel prices: dominant firms, local monopolies and ‘captive’demand 411 REGIONAL STUDIES, REGIONAL SCIENCE