Gasoline Pricing in the Country and the City


In many markets, prices adjust quickly when costs rise, yet adjust sluggishly when costs fall. Such asymmetric pricing has received particular attention in retail gasoline where larger asymmetry has been related to greater market power. Using novel data from urban and rural gasoline markets, I document new evidence on this relationship by providing the first estimates of asymmetric pricing from rural towns. I find substantial asymmetry in these markets, show that local market concentration is positively related to asymmetry, and highlight the potential role of role collusion in generating these patterns.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2
Fig. 3


  1. 1.

    For example, see “Oil’s plunging—why hasn’t gasoline fallen faster?” from

  2. 2.

    During the sample period, prices were also uploaded to a (now defunct) province-specific website ( as well as city-specific websites (e.g.,

  3. 3.

    GasBuddy is North America’s most popular gasoline price website (Lewis and Marvel 2011).

  4. 4.

    Station-level price observations are less than the product of total days and stations (\(1947 \times 378= 735,966\)). This is because there are days where some stations do not have GasBuddy price reports. All markets have price data for at least 250 days.

  5. 5.

    See Statistics Canada’s Geosuite tool and related documentation for details on how CMA’s, CA’s, and RAs are defined:

  6. 6.

    Given that the GasBuddy data are self-reported, sample representativeness is a potential issue. In Section A.1 of the online appendix, I validate the GasBuddy data using independently-collected market-level data on gasoline prices from MJ Ervin & Associates.

  7. 7.

    These market structure variables are based on simple station counts. They are not based on weighted-station counts, where each station count is weighted by its relative volume of gasoline sold. Unfortunately, station-level volume data are unavailable for constructing such weights.

  8. 8.

    MJ Ervin collects station-level rack price data from a sample of non-branded gasoline companies from many local markets in Canada. It then computes the daily rack price for a given market as the average wholesale price that is faced by all stations that are in that market.

  9. 9.

    Based on the 2007 MJ Ervin report, the major refiners and wholesale distributors of gasoline in Ontario in 2007-08 are: Suncor/Sunoco (1 refinery, 6 terminals, 1 bulk plant in Ontario); Imperial Oil/Esso (2 refineries, 7 terminals, 21 bulk plants); Shell (1 refineries, 7 terminals, 0 bulk plants); and Petro-Canada (1 refinery in nearby Montreal, 5 terminals, 43 bulk plants). These four firms are vertically integrated: They compete in both the upstream and downstream markets for gasoline. They use their distribution network of terminals and bulk plants across Ontario to supply wholesale gasoline to the vast majority of stations in the province: both branded and independent. Data on retailers’ wholesale supply contracts are unfortunately unavailable.

  10. 10.

    For cities, the measure of daily highway traffic that drives past a market is primarily based on highway segments from Ontario’s major 400-series highway network. This highway runs past most major cities in the province. For smaller cities and towns, the measure is primarily based on traffic flows from highway segments on the Trans-Canada highway. For rural markets that are not on the Trans-Canada highway, their measure is primarily based on secondary highway road segments. Regardless of the highway type that is used to measure average daily highway traffic, the measures are all based on the same metric: average daily vehicle counts. The measure is therefore comparable across markets and is useful for capturing cross-market differences in the volume of highway traffic that drives past a market day-to-day.

  11. 11.

    Ervin & Associates (2010) estimates that stations in urban markets, such as Toronto, pump 7–8 million liters of fuel per year, whereas rural stations pump only 1–2 million liters.

  12. 12.

    The differences in the rack price series across panels A–C of Fig. 1 reflect differences in the nearest rack price locations for Ajax, St. Thomas, and Atikokan. The rack price series for these markets come from Toronto, Hamilton, and Thunder Bay, respectively. See Sect. 2.1.2 above for the list of rack price locations, and how local retail markets are matched to them.

  13. 13.

    Throughout, I study unweighted average prices across stations in a market. I am unable to compute volume-weighted averages, as daily station-level volume data are unavailable.

  14. 14.

    My use of weekly frequencies contrasts with previous studies on pricing asymmetry at daily frequencies in urban markets (e.g., Lewis and Noel 2011). Given that urban markets have daily wholesale deliveries, the use of daily price data in these studies is appropriate. As a check, I have estimated asymmetric pass-through models at daily frequencies for my sample’s urban markets (e.g., CMAs). Inferences regarding cost pass-through in these markets (rapid, symmetric pass-through) are similar if I use daily or weekly data.

  15. 15.

    See Lewis and Noel (2011) for an analysis of asymmetric pass-through for cycling markets that uses regime switching regressions to estimate daily rates of pass-through asymmetry.

  16. 16.

    Panel A of Table A.1 in the online appendix presents Chow-test results that show that it is appropriate to estimate separate models for cost-based and sticky markets rather than estimating a single pooled model for these market types. The rate of cost pass-through and pricing asymmetry for cycling markets from Table 1 sits between sticky and cost based markets. These are omitted for the sake of brevity, and to keep the analysis focused on my novel results with regard to pricing asymmetry in rural markets.

  17. 17.

    The urban markets exhibit a remarkable lack of asymmetry in price adjustment relative to what is found in previous studies (e.g., Lewis and Noel 2011). This may be explained by findings from Atkinson et al. (2014): They show that at the start of 2007 (6 months before my sample starts), gasoline pricing permanently shifts from cycles to cost-based pricing in Ontario’s urban markets. This shift occurs just days after a wholesale supply shock: a major fire at Imperial Oil’s Nanticoke refinery. The authors provide evidence from Toronto that this shift to daily cost-based pricing corresponds to a shift to within-day price cycles at even higher frequencies. For instance, prices are high in the morning and low in the evening, and hence are constantly adjusting to daily cost shocks that stem from daily wholesale fuel deliveries. If this shift in pricing is unique to Ontario’s urban markets, this could explain the lack of pricing asymmetry in my sample’s urban markets relative to what has been previously found.

  18. 18.

    Chow test results in Panel B of Table A.1 in the online appendix support using separate price response regressions for positive and negative cost shocks rather than using pooled regressions.

  19. 19.

    Let \(\beta _{i0}^{+}, \beta _{i0}^{-}, \beta _{i1}^{+}, \beta _{i1}^{-}, \gamma _{i1}^{+}, \gamma _{i1}^{-}, \phi _{i1}^{+}, \phi _{i1}^{-}\) and \(\psi _{i}\) be market i’s error correction model parameters from (1a) and (1b). For \(\tau =0\), for positive shocks \(\varDelta p_{i0}=\beta _{i0}^{+}/2.5\), and for negative shocks \(\varDelta p_{i0}=\beta _{i0}^{-}/(-2.5)\). For \(\tau =1\), for positive shocks \(\varDelta p_{i1}=\Big (\beta _{i0}^{+}+\beta _{i1}^{+}+\max \big (\gamma _{i1}^{+}\beta _{i0}^{+},0\big )+\min \big (\gamma _{i1}^{-}\beta _{i0}^{+},0\big )+\max \big (\phi _{i1}^{+}\times (\beta _{i0}^{+}-\psi _{i}),0\big )+\min \big (\phi _{i1}^{-}\times (\beta _{i0}^{+}-\psi _{i}),0\big )\Big )\Big /2.5\), and for negative shocks \(\varDelta p_{i1}=\Big (\beta _{i0}^{-}+\beta _{i1}^{-}+\max \big (\gamma _{i1}^{+}\beta _{i0}^{-},0\big )+\min \big (\gamma _{i1}^{-}\beta _{i0}^{-},0\big )+\max \big (\phi _{i1}^{+}\times (\beta _{i0}^{-}-\psi _{i}),0\big )+\min \big (\phi _{i1}^{-}\times (\beta _{i0}^{-}-\psi _{i}),0\big )\Big )\Big /(-2.5).\)

  20. 20.

    Regressions for two or more weeks following a cost shock yield statistically insignificant pass-through asymmetries across markets with different numbers of firms and remoteness from wholesale supply. This aligns with Fig. 3, which shows that asymmetry does not persist in rural markets 2 weeks after cost shocks occur.

  21. 21.

    The excluded categories in the regressions are > 25 Stations and < 100 km from wholesale supply: Urban markets that are close to wholesale supply.

  22. 22.

    Consider a market with four stations: two operate under Shell’s brand name; one operates as Esso; and one operates as Pioneer (an independent). In this case, the two Shell stations would have 50% brand shares. The Esso and Pioneer stations would each be assigned 25% brand shares.

  23. 23.

    Unfortunately, data on additional station-level characteristics such as number of pumps, size of gasoline storage tanks, vertical contracts, and convenience store presence are unavailable.

  24. 24.

    The Chow-test results in Panel B of Table A.1 in the online appendix support the use of separate regressions for positive and negative shocks for all oil major/independent and urban/rural subsamples instead of pooled regressions. Moreover, the Chow-test results in Panel C of Table A.1 support the use of separate pass-through regressions for major and independent stations, and for urban and rural markets.

  25. 25.

    See for details on these price-fixing cases.


  1. Atkinson, B., Eckert, A., & West, D. S. (2014). Daily price cycles and constant margins: Recent events in Canadian gasoline retailing. The Energy Journal, 35(3), 47–69.

    Article  Google Scholar 

  2. Bachmeier, L., & Griffin, J. (2003). New evidence on asymmetric gasoline price responses. Review of Economics and Statistics, 85(3), 772–776.

    Article  Google Scholar 

  3. Bacon, R. W. (1991). Rockets and feathers: The asymmetric speed of adjustment of UK retail gasoline prices to cost changes. Energy Economics, 13(2), 211–218.

    Article  Google Scholar 

  4. Balmaceda, F., & Soruco, P. (2008). Asymmetric dynamic pricing in a local gasoline retail market. Journal of Industrial Economics, 56(3), 629–653.

    Article  Google Scholar 

  5. Borenstein, S., Cameron, C., & Gilbert, R. (1997). Do gasoline prices respond asymmetrically to crude oil price changes? Quarterly Journal of Economics, 112(1), 305–339.

    Article  Google Scholar 

  6. Byrne, D. P., Leslie, G. W., & Ware, R. (2015). How do consumers respond to gasoline price cycles? The Energy Journal, 36(1), 115–147.

    Article  Google Scholar 

  7. Chandra, A., & Tappata, M. (2011). Consumer search and dynamic price dispersion: An application to gasoline markets. RAND Journal of Economics, 42(4), 681–704.

    Article  Google Scholar 

  8. Clark, R., & Houde, J.-F. (2013). Collusion with asymmetric retailers: Evidence from a gasoline price-fixing case. American Economic Journal: Microeconomics, 5(3), 97–123.

    Google Scholar 

  9. Deltas, G. (2008). Retail gasoline price dynamics and local market power. Journal of Industrial Economics, 56(3), 613–628.

    Article  Google Scholar 

  10. Eckert, A. (2002). Retail price cycles and response asymmetry. Canadian Journal of Economics, 35(1), 52–77.

    Article  Google Scholar 

  11. Houde, J.-F. (2012). Spatial differentiation and vertical mergers in retail markets for gasoline. American Economic Review, 102(5), 2147–2182.

    Article  Google Scholar 

  12. Lewis, M. S. (2011). Asymmetric price adjustment and consumer search: An examination of the retail gasoline market. Journal of Economics and Management Strategy, 20(2), 409–449.

    Article  Google Scholar 

  13. Lewis, M. S., & Marvel, H. P. (2011). When do consumers search? Journal of Industrial Economics, 59(3), 457–483.

    Article  Google Scholar 

  14. Lewis, M. S., & Noel, M. D. (2011). The speed of gasoline price response in markets with and without Edgeworth cycles. Review of Economics and Statistics, 93(2), 672–682.

    Article  Google Scholar 

  15. Ervin & Associates, MJ. (2010). National retail petroleum site census

  16. 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.

    Article  Google Scholar 

  17. Peltzman, S. (2000). Prices rise faster than they fall. Journal of Political Economy, 108(3), 466–502.

    Article  Google Scholar 

  18. Tappata, M. (2009). Rockets and feathers: Understand asymmetric pricing. RAND Journal of Economics, 40(4), 466–502.

    Article  Google Scholar 

  19. Verlinda, J. A. (2008). Do rockets rise faster and feathers fall slower in an atmosphere of local market power? Evidence from the retail gasoline market. Journal of Industrial Economics, 56(3), 581–612.

    Article  Google Scholar 

  20. Yang, H., & Ye, L. (2008). Search with learning: Understanding asymmetric price adjustments. RAND Journal of Economics, 39(2), 547–564.

    Article  Google Scholar 

Download references


I am grateful to Dustin Coupal of GasBuddy Organization Inc. for providing the data used in this study. Ingrid Burfurd, Charlie Shenton, and Daniel Tiong have provided excellent research assistance. The two anonymous referees, numerous seminar participants, and especially the editor, Lawrence White, have provided many helpful comments.

Author information



Corresponding author

Correspondence to David P. Byrne.

Additional information

Funding from the University of Melbourne FBE Faculty Research Grant scheme is acknowledged. The views and opinions expressed in this article are solely those of the author. All links referenced in footnotes were current as of 25/8/2018. The paper’s online appendix can be found at:

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 856 KB)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Byrne, D.P. Gasoline Pricing in the Country and the City. Rev Ind Organ 55, 209–235 (2019).

Download citation


  • Asymmetric pricing
  • Cost-based pricing
  • Retail gasoline
  • Sticky pricing

JEL Classification

  • L11
  • L9
  • D22