The Post Which Claims California Will Have $8/gallon Gasoline is Factually Incorrect and Betrays a Fundamental Lack of Understanding of Energy Markets.
And it's actively harmful to the policy debate about how we responsibly wind down fossil fuel infrastructure
In April, Valero announced its planned closure of its Benicia refinery (henceforth “Benicia”). Benicia produces roughly 75,000 barrels of gasoline/day, or roughly 8.5 percent of California’s daily demand. The closure—which will occur in 2026–was met with concern from policymakers in the state because not only was it unexpected, but it was the second such closure announced in the past year after Phillips 66 announced it was closing its Los Angeles refinery at the end of 2025. Together, these two closures mean that almost 20 percent of CA gasoline refining capacity will come offline within two years.
These events have understandably led to concern about the future of gasoline prices in the state. Even if California does make good on its plan to completely stop selling new gas-powered cars by 2035, spikes in gasoline prices for the foreseeable future will still be important for California consumers and firms. After all, California’s gasoline market is larger than $60B/year, meaning that a persistent spike in gasoline prices translates to billions every year that come out of people’s paychecks. In response, policymakers are evaluating what impacts these closures have on a variety of dimensions. Most important for our purposes here is the price paid at the pump by California consumers and firms.
And this is what brings us to the post put online by Michael Mische, who is a consultant and Professor of the Practice at USC’s Marshall School of Business, an affiliation he claims directly at the top of the post. The main result from Mische’s post that’s been circulating widely among the media (including in the Wall Street Journal) is that this reduction in refining capacity could result in California having $8/gallon gasoline within two years. As I’ll detail in this post, this result is not credible. It’s based on a poor grasp of basic economic concepts, a faulty understanding of California’s gasoline market, extreme assumptions with no theoretical or empirical backing, and several factual errors.
Misunderstanding of basic economic concepts
The post begins with the statement: “California can ill afford the loss of one refinery, let alone two.” There is no reason why this is per se true. In the U.S., only 30 states have at least one refinery. Which means 20 states—that collectively hold roughly 40 percent of the population–have precisely zero refining capacity and rely 100 percent on imports. As I detail below, if California refineries close over an extended period of time, giving global markets time to adapt to serving the California market, then the impact of refinery closures on consumers would be minimal.
Mische proceeds to fundamentally misunderstand a crucial, but basic, economic concept. He claims that California has historically relied on refineries in Washington State to fill its production gaps, and when that import supply is maxed out, the state must rely on Gulf Coast and foreign refineries to meet demand. This is mostly correct.[1] But then Mische claims that “[A]s a consequence of the two refinery closings, California will be at the mercy of out of state and foreign, non-U.S. refiners.”
And this is the basic economics that Mische misses, which (mis)informs the rest of his piece. Gasoline is a commodity that is traded on a global market with robust international competition.[2] What this means in practice is that the price of (wholesale) gasoline is set by the marginal barrel, which is the highest cost barrel to bring to market.[3] Because California is already a net importer of crude, gasoline, and gasoline blending components, it already is “at the mercy” of out-of-state production. That is, the price currently reflects the fact that the last unit brought to market comes from a refiner outside of California. Additional refinery closures of course reinforce this dynamic, but do not change it.[4]
To be clear, additional refinery closures do present policy challenges that will require a deft hand, and close coordination, from both industry and the State. Refineries in California provide between 6,000-7,000 high-paying, union jobs. They are additionally important tax bases for their local communities. But they also are a source of substantial local environmental pollution. Furthermore, relying more heavily on imports will require creating more port capacity to receive imports. This can be done by converting defunct refineries to product terminals, as has been done at the Rodeo and Martinez refineries, but will likely need the state to streamline permitting to do so in a timely manner. But it does not mean that gasoline supply will cataclysmically shrink, resulting in large price spikes. If anything, the opposite is true—conditional on the state facilitating an increase in import capacity, first-principles tell us that the price impact should be negligible.
Inconsistent and hidden analytical framework
Mische states that “multiple models indicate that the pending loss of two refineries and a corresponding reduction in gasoline production will result in higher retail prices.” Mische gives extremely limited information about what these “models” entail. It is standard practice in public policy analysis—and especially in academic economics—to provide detailed information about the model(s) an analyst is using in order to inform the debate. Mische eschews this practice entirely and simply reports results.
Of the few details Mische does give, two are extraordinary assumptions that are not guided by any economic theory or data. The first is that Mische assumes that of the lost refining capacity, only 20 percent is replaced by imports from Washington State. The second is that “[D]emand (consumption) was maintained at a static level,” or in economic terms, Mische assumes the (long-run) price elasticity of demand and supply is zero.
In practice, these two assumptions mean that under Mische’s models, if prices increase: 1) no refiner in the world (beyond those in Washington State) would respond to higher gasoline prices in California by producing more CARBOB (wholesale gasoline sold in California), even if the profit margins of doing so were astronomical and 2) consumers would buy the same quantity of gasoline at any price.[5] The implications of these assumptions, applied in the extreme, would mean that if gasoline prices were $50/gallon, California consumers would still consume 13.2 billion gallons of gasoline annually, spending over $660 billion, or more than 1/3rd of all the wages and salaries earned in the state.
Basic factual errors
Mische’s post is also filled with basic factual errors. He writes that “[A]t no time has California ever faced a permanent 20% reduction in gasoline production.” This is very wrong. The figure below demonstrates that from 1982-1995, California’s gasoline refining capacity declined by about 1/3rd, while at the same time gasoline demand increased by 25 percent.[6] Over this period, however, the real (inflation-adjusted) price of gasoline decreased.
As the figure shows, though, the main difference between then and now is that California previously had a surplus of gasoline, whereas now the State consistently operates with a deficit. Regardless, shrinking capacity does not automatically mean that prices will spike, which is a constant assumption applied without economic justification throughout Mische’s piece.
Mische also claims that California regulations are to blame for higher gasoline prices. In Exhibit 5.0, Mische lists a variety of regulations and then compares gasoline price changes in California to the rest of the U.S. since the inception of the regulation. This is not an economically meaningful analysis or way to analyze regulatory impacts on prices. For example, he shows gas prices before and after the State passed ABX2-1 in October 2024.[7] But you would be hard-pressed to find a credible industry analyst or economist who believes the regulation (which simply gave the California Energy Commission the authority to create minimum inventory levels of gasoline) is responsible for gasoline price movements. Instead, the proximate cause of gasoline price movements in the short-run is both local (e.g., refinery fires) and global (e.g., changes in the price of crude oil).
In Exhibit 7.0, Mische shows his “estimates” of the impact of regulatory fees on gasoline prices in the State. As he does throughout the piece, he provides no description of how he arrived at these estimates. Among these, Mische states that as of April 2025, the cost of cap and trade and “other environmental programs” (presumably, the Low Carbon Fuel Standard, or LCFS) is 55 cents/gallon, that the “CA Special Blend” (presumably, the additional refining cost of making CARBOB) is 15 cents/gallon, and that the “CA Seasonal Blend” (presumably, low-Reid Vapor Pressure (RVP) summer CARBOB that’s required during summer months) adds another 12 cents/gallon. Two of these are wrong, and one of them is highly debated.
First, the Oil Price Information Service (OPIS) computes the cost of California environmental programs on gasoline and finds it to be roughly 40 cents/gallon during the period he selects—15 cents/gallon below Mische’s estimate. Second, the incremental cost of blending CARBOB (relative to RBOB, the wholesale gasoline the rest of the country uses in summer months) has likely fallen over time as non-California standards on gasoline pollutants have tightened and are now estimated by several analysts to be no greater than 10 cents/gallon.[8] There is robust debate about this number, though, but that is not acknowledged in any form by Mische. Finally, Mische’s “seasonal blend” estimate of 12 cents/gallon is a reasonable estimate for the additional cost of low-Reid Vapor Pressure (RVP) summer gasoline, but it is incorrectly applied to both the summer and winter. In the winter, higher RVP gasoline is sold, which is cheaper to produce. Taken together, these estimates incorrectly add roughly 25 cents/gallon to the average gasoline price.
Poorly constructed policy remedies
The conclusions Mische draws from his mystery models lead him to recommend a host of policy changes. Most troubling of all is his first recommendation, which is to essentially subsidize the closing refineries to stay in business. As my Stanford colleague Neale Mahoney and I have written in a separate analysis examining the logistics and price impacts associated with Benicia’s closure, giving operating subsidies “raises two concerns. First, asymmetric information between the refiners and the state creates significant overpayment risk; the refiners know more about their operating costs than the state. If the state opts to subsidize, the refiners will push for the largest possible subsidy, and the state may not have the information or risk appetite to push back. Second, if the state does subsidize one of the refiners, it may trigger a vicious cycle in which other financially stable refineries threaten to shut down in order to receive their own subsidies.”
The rest of Mische’s recommendations are essentially all calls for repealing existing regulation or reducing fuel taxes. While it is certainly worthwhile for policymakers to constantly evaluate the efficacy and usefulness of regulations, nowhere does Mische actually discuss the benefits of any of the regulations that he calls to abolish or taxes he calls to lower. For example, he calls for the gasoline excise tax to be lowered to the national average and for the cap and trade program to be “capitated” at $0.65/gallon. To be sure, following through on some of these recommendations would reduce fuel prices, particularly lowering the excise tax. But they would also increase pollution in a state that already has some of the worst air pollution in the country. Standard climate-economics frameworks attempt to quantify these harms, and therefore a responsible analysis would examine the dollar trade-offs between the direct benefit of lower fuel prices and the indirect costs of worse health outcomes resulting from pollution.
Unhelpful for the policy debate
Managing the decline of fossil fuel consumption and production as green technologies come online was always going to be a challenging task, and the state needs all the rigorous, well-structured analysis it can get its hands on. Unfortunately, this post by Mische is not that. Among other things, the post does not analyze the economics of California’s gasoline market correctly; it overestimates the costs of California’s regulations, and fails to consider their benefits; and it makes assumptions, such as perfectly inelastic demand and supply for gasoline in California, that are not scientific and lead to fantastical conclusions. His lack of methodological transparency and rigor ultimately makes it difficult for other academic and policy experts to engage with his ideas. And, the attention-grabbing headline of $8/gallon gas could harm public discourse by pressuring policymakers to make bad decisions, such as subsidizing the Benicia refinery. Ultimately, policymakers should not view the report as credible, rigorous, or useful.
[1] Mische states that the Asian refineries that are supplying CA with imports include “South Korea and China.” The main Asian countries that provide import to supply to California are India and South Korea, which using EIA data, I’ve calculated that since 2010, India and Korea have supplied California with 30% of its gasoline (and gasoline blending components), while China has only supplied 2%. In other words, China is not a crucial swing supplier to California gasoline markets.
[2] For ease of exposition, I’m using finished gasoline and gasoline blending components—the ingredients that, when combined, create finished gasoline—interchangeably. When evaluating imports into California, it is common practice to group both finished gasoline and blending components together.
[3] There are of course caveats to this. For example, in the short-run, prices can trade above or below marginal cost depending on inventories or if there’s collusion (such as is the case with OPEC and crude oil). But Mische is focusing his analysis on the medium-to-long-run, in which case economists widely believe that prices are set by the highest marginal cost producer.
[4] With that being said, there are important short-run vs. long-run dynamics to consider (which again, Mische does not). In the short-run, we can think of the supply curve of CARBOB (wholesale gasoline sold in California) as upward sloping, meaning every additional unit to bring to market is more expensive than the last. This is because very few, if any, global refineries are committed to making CARBOB full time. For them to switch, an arbitrage needs to emerge between global and California prices. In the long-run, however, refinery closures in California will present a greater opportunity to penetrate one of the world’s largest gasoline markets, and we should expect both global CARBOB capacity and actual supply to increase, gradually eliminating short-run arbitrages.
[5] To make matters even more confusing, it’s not clear how the situation Mische modeled is even possible. Mische assumes roughly 9.5 million gallons/day of gasoline production comes offline, then adds back roughly 2 million gallons/day to account for predicted Washington State imports (but does not include other, foreign imports). This leaves a gap of 7.5 million gallons/day between supply and demand. In any standard economic model, prices would clear the market, which would in lower quantity demanded. But again, Mische assumes that the quantity demanded is constant, thereby creating a gap which, by construction, cannot be filled.
[6] As noted in the graph, I am roughly estimating gasoline refining capacity here by assuming the yield of gasoline has held constant over time, as has the share devoted to in-state production. Because these assumptions are both simply scalars, they do not affect the overall percentage change in capacity (i.e., the percent change is the same as taking total refining capacity change).
[7] I, along with my Stanford colleague Neale Mahoney, publicly supported this regulation based on the economic problems the regulation sought to address. See https://www.sacbee.com/opinion/op-ed/article293111404.html
[8] Specifically, by 2016 the average gallon of gasoline sold in the US was already satisfying CARFG’s 1) benzene, 2) oxygen, and 3) aromatics requirements (the Sulfur requirements are likewise very close). See, e.g., EPA Fuel Trends Report for national standards and CARFG standards for CA standards.1