Widespread adoption of electric vehicles—combined with a shift in travel modes towards more walking, cycling, and transit use—can help ease the climate crisis, improve quality of life, and save Americans money. A key to shifting travel modes to less automobile use is making biking, walking, and transit safer and more convenient by redirecting infrastructure investments and making urban areas more compact. When compared to the trajectory we are currently on, faster vehicle electrification and a shift in travel modes could save the US economy a cumulative $13 trillion, save the average urban resident $2,000 per year, reduce transportation inequities, and make cities more livable.

A recent study, from ITS-Davis and the Institute for Transportation & Development Policy, explores four scenarios and identifies steps that can move us in the direction of cost savings and lower emissions of greenhouse gasses and pollutants:  

• In the Business as Usual scenario, vehicle electrification and a shift in travel modes occur gradually, according to current trends and policies.  
• In the Electrification scenario, the transition to electric vehicles is faster – as fast as experts consider possible.  
• In the Mode Shift scenario, people use non-automobile travel modes at a “maximum feasible” level as these modes become more feasible and appealing. 
• In the fourth Electrification + Shift scenario, these last two scenarios are combined.  

Electrification 

The Electrification scenario has sales of new light-duty vehicles (cars and light trucks) reaching 60% electric nationwide by 2030 and 100% by 2050. The US reached about a 7.6% electric (and plug-in hybrid) vehicle market share in 2023, which is a good start, but there is a ways to go.

Achieving the Electrification scenario will probably require that eventually: 1) electric vehicles provide significant ownership savings compared to gasoline vehicles; 2) electric vehicle ranges are sufficient to meet all driving situation needs, and 3) public charging becomes widely available at or close to most homes and in public roadside locations convenient for long trips. If these conditions can be met, then the main challenge will be a marketing one: convincing Americans that EVs are desirable and meet their needs.

Mode Shift

Getting Americans to reduce their dependence on cars and use other modes of transportation may seem harder than going electric, though the changes envisioned in our Mode Shift scenario are not massive. In this and the combined Electrification + Shift scenario, urban car, SUV, and personal pick-up truck (light-duty vehicle) travel in 2050 would be about 25% less than in the Business as Usual scenario. This reduction goes along with much better infrastructure to increase travel by transit, biking, and walking. And even with the increases in infrastructure and operations for these modes, the shift will save governments and individuals considerable funds and provide an important range of health benefits, which are described in the report.

But what changes are needed to make such a mode shift possible? It will require increasing density and mixed-use development in cities and suburbs, coupled with more sidewalks, protected bike lanes, and safer streets with slower traffic. All of this will encourage people to choose to walk or ride a traditional or electric bike or an e-scooter. Electric bikes could play a growing role for trips under 5 miles if roadways are designed to make this safe. On the transit side, there is a critical need for more frequent service, which could be made possible by increased population density and would make using transit more appealing and logical. People will need to consider these modes convenient and safe and at least give them a try. Having the right infrastructure and shorter trips is a prerequisite. 

American cities are already getting denser, but the trend is gradual. The average neighborhood has about 12,500 people per square mile today, and we’re on track to increase to 13,500 by 2050. But achieving the Mode Shift and combined scenario will require reaching an average density of nearly 17,000 people per square mile by 2050. This doesn’t mean that everywhere will look like New York City—it just means that the average American city will look more like Los Angeles than Atlanta, the average neighborhood more like Arlington, Virginia than Arlington, Texas.  

Small and large cities will have a density and layout more resembling a walkable town than spread out areas of housing, miles from shopping plazas. Such moderate densities can be achieved through public policy that is already being adopted in some states. Examples include “missing middle” policies to legalize multi-family housing up to six or eight units on all lots and removal of requirements for a minimum of off-street parking. The densified cities in the mode shift scenarios will not require anyone to leave or redevelop their home against their wishes—it will only create a supply of walkable neighborhoods that are currently in high demand. 

The expansion of public transit and protected bike and pedestrian ways will require investment, but the funds required are less than what is currently spent on building and maintaining road infrastructure. The estimated cumulative savings for national, state, and local governments between 2024 and 2050 would be $2 trillion. Similarly, much of the annual savings of $2000 for urban residents would come from the reduction in car ownership.  

Action Plan

The most important point is that we need citizens and governments committed to making this happen, voting for it, and re-allocating sufficient funds. The report on our study provides a list of policy strategies at federal, state, and local levels that would enable these changes. Some examples include:  

• Modestly increase the tax on internal combustion engine cars to help fund incentives to lower the costs of electric vehicles and to fund transit and infrastructure for cycling and walking. 
• Support equitable placement of public charging points and charging in multi-unit dwellings. 
• Reallocate federal and state transportation budgets and road space to walking, cycling, and public transport. 
• Entirely stop building new roads or expanding existing ones. Use the funding to maintain and improve the capacity of existing roads by reallocating space to more efficient modes, like cycling and public transport. 
• Along major roads, build a connected, integrated network of bus rapid transit lines—buses that operate like light rail: have a dedicated center-running lane, take priority at intersections, and are boarded from stations. 
• Redirect existing subsidies for fossil fuels to the expansion and decarbonization of the nation’s electricity system to support clean transportation and other uses. 

This transformation is possible. Cities around the country have already proven that it can be done. In only two years, Seattle transformed its bus network such that 64% of residents live near a bus line with ten-minute service or better, up from only 25% before the change, leading to increased ridership. Arlington, Virginia demonstrated the importance of rapid transit: by routing a Metro line through commercial neighborhoods and building density around the stations. Arlington was able to dramatically increase population and tax revenue without any increase in traffic. In the future, cities around the country could replicate this success using affordable bus rapid transit instead of rail. The state of Minnesota is debating an act that could remove parking minimums all across the state, cutting required parking lot construction, and thus making additional space available for housing, shops, and walking.  

In 30 years, our cities could be greener, more walkable, more efficient, less congested, quieter, and more affordable. We could save money for the public and private sectors. We could reduce air pollution, social segregation, and the cost of living. All that’s needed is the collective willingness to act.  

——————– 

Lewis Fulton is Director of the Energy Futures Program at ITS-Davis. 

In an effort to curb emissions, governments in major vehicle markets are proposing and adopting requirements that electric vehicles (EVs) make up a certain percentage of new vehicle sales in coming years. This week the Government of Canada announced EV sales targets of 60% by 2030 and 100% by 2035 for light-duty vehicles, similar to the goals the United States announced last year of 50% by 2030 and 67% by 2032. Other countries and regions such as China, Japan, and Europe have also made commitments to specific targets.

These targets raise a key question: Can enough electric vehicles be produced in time across the range of countries making these commitments?

Answer: Maybe. There is a great deal of uncertainty, but this is clear: keeping trade open will help, and investing in domestic EV production is critical.

In a recent study, we modeled scenarios in which governments in six electric vehicle markets—the US, Europe, Canada, Mexico, Japan, and South Korea—adopt sales targets requiring EVs to make up a certain percentage of their new vehicle sales by 2030. Depending upon whether these governments adopt higher or lower proposed targets, the markets will need 12 to 25 million new EVs per year by 2030.

The chart below shows the comparisons between higher and lower national EV sales targets, on one hand, and higher and lower company planned EV production capacity, on the other hand. The higher and lower planned production capacities represent “tentative” and “firm” plans and investments, based on our review of related announcements by automakers beginning in 2020.

As shown by the right-hand bar in the chart, if these six governments were to adopt the lower proposed targets, their combined EV production capacity, according to “firm” plans, would be just enough for the required 12 million EV mark. But, if they adopt higher targets (here assumed to be 50% EV sales shares by 2030), their combined production capacity could have a shortfall of 9 to 13 million EVs in that year—which is a third to half of the total number of EVs needed across the countries.

If national shortfalls are relatively small and limited to few of these individual markets, then trade among them could make the combined goals achievable. For example, the US and Europe—the two largest EV markets examined—can achieve their lower sales targets if their planned capacities are pooled. So transatlantic trade may serve as an important approach for reaching combined targets.

On the other hand, with the higher targets, especially combined with lower “firm” production capacities, the shortfalls are too great to be made up only through trade among these countries. In the US-Europe example, their combined annual production shortfall would reach 10.6 million EVs in 2030. This may be addressed with additional automaker investment in each region; if done in line with each region’s needs, the total investment would be $108 billion, which would require an additional $66 billion above the current firm commitment of $42 billion. In this case trade—between these countries and also including additional countries such as China and India—can still play a key role, since it would provide more flexibility regarding where these needed investments can be made.

However, while trade of EVs and their parts can help, trade dynamics are complicated and subject to limitations by regional and national policies and geopolitical matters. These limitations reinforce the need to boost investments in production capacity within each market, though it is unclear that each market will be able to meet 100% of its own demand in the high target scenarios. We believe it is very important to keep trade options as flexible as possible.

Another factor affecting supply is that manufacturers are understandably wary of making large EV production investments in the short-term, because producing the required number of EVs to meet a sales target does not guarantee that consumer demand will follow and protect the companies from financial loss. So attracting the large investments needed is complicated, and policies and incentives that support demand are key. One hopeful example is California, which has an EV sales target of 100% by 2035 and policies supporting demand. In that state, EV sales have skyrocketed from about 7% of new vehicles in 2020 to nearly 27% in 2023.

We do not know yet how targets, demand, or supply growth will play out. Other unknowns not examined or modeled in our study, such as availability of raw materials for batteries and batteries themselves, may also affect production capacity. Right now, the situation looks challenging in the 2030 timeframe. The US and other countries may need to encourage greater investments in EV assembly capacity so this can ramp-up more quickly. More trade may also help, but comes with its own limits and risks.

—

Hong Yang is a PhD candidate in the UC Davis Energy Systems Program.

Lewis Fulton is the director of the Energy Futures Program at ITS-Davis and the former director of the Sustainable Transportation Energy Pathways Program at ITS-Davis.

On the path to global transportation sustainability, electric vehicles (EVs) are making inroads, but larger cars with bigger carbon footprints are hampering climate progress. The right policies can correct our course.

A recent report by the ITS-Davis European Transport and Energy Research Centre, the FIA Foundation, and the Global Fuel Economy Initiative identifies trends in the global car market from 2010 to 2022 and their impacts on energy consumption and CO2 emissions. The report highlights increasing light-duty EV sales that have substantially reduced energy consumption and CO2 emissions alongside rising SUV sales that have counteracted these reductions.

Two Steps Forward, One Step Back

Energy and direct CO2 emissions intensities for new vehicles declined globally at a rate close to 2% per year between 2005 and 2022. From 2020 to 2022, average energy consumed per mile for new light-duty vehicles decreased faster, by 4.2% annually, on average. CO2 emissions per mile declined almost 6% annually in the same period. These accelerations are mainly due to an increase in EV sales.

Emissions could have fallen farther, however.

A persistent shift to larger and heavier SUVs is offsetting benefits from EVs. Global sales of SUVs increased from 22% of all light-duty vehicles in 2005 to 51% in 2022. The case of the US is extreme, by global standards, with SUVs and pick-up trucks now accounting for 75% of all vehicles sold.

If the global average weight of vehicles had stayed constant since 2010, energy and CO2 emission intensities of combustion engine vehicles would have declined about 30% faster. The shift to SUVs also exacerbated equity and road safety challenges, due to their higher prices and danger to other road users.

This trend toward bigger and heavier vehicles is not limited to combustion engine vehicles. EVs are also getting larger, creating more demand for critical materials, such as lithium and cobalt, that are subject to supply risk—adding to possible bottlenecks for EV production in the absence of policy action. The report on vehicle trends offers policy suggestions to steer markets toward cleaner vehicles and away from large and heavy vehicles.

Better policy needed to meet emissions goals

Policies encouraging the downsizing of vehicles and boosting of the EV market are needed to reduce climate change and meet net-zero greenhouse gas emissions goals.

Downsizing (and down-weighting) vehicles would also increase road safety and reduce inequities, especially in countries that have limited alternatives to car travel. Downsizing would make EVs more affordable and reduce the amount of critical minerals needed for batteries, thereby reducing the social, environmental, and geopolitical security risks from mining and mineral processing.

Policy recommendations

• Reduce the sales-weighted average footprint and weight of vehicles over time.
• Reduce battery weight and capacity (kilowatt-hours per vehicle) using a corporate average approach, as is done with fuel economy standards. These limits would restrict the growing demand for critical materials and the negative human and environmental impacts from mining, including in low-income and underserved communities.
• Regulations on the size and weight of vehicles (including EVs) should reward innovation (and reduce costs) by allowing the trading of credits across vehicle categories and companies, as is now done in Europe and the US with emissions and fuel economy standards.
• Eliminate regulatory carve-outs for SUVs and pickup trucks—which have been used in the US and are a primary factor in the rapid transition from cars to SUVs.
• Introduce more stringent environmental regulatory requirements and require more stringent safety standards for heavily-used vehicles, such as those used by ride-hailing and taxi drivers since the EV transition is more cost-effective in these cases, and road safety benefits are higher for vehicles that travel more.
• Regulations and incentives should be expanded to allow universal and equitable access to EV charging infrastructure—including requiring or funding charger installations in or near multi-unit dwellings.
• Vehicle taxes should be differentiated on the basis of energy efficiency and environmental impacts. They should also be modulated based on size, weight, and price to respond to challenges for equity, road safety, and material demand. Taxes and charges adopted at the local level can also be tailored to respond to these effects. Key examples already exist in Paris and other French cities, which will start applying in 2024 parking fees differentiated based on size and weight and engine type.

A shift away from SUVs, paired with tailored and continuous support for EVs, will be crucial to reducing CO2 emissions, enhancing energy savings, improving road safety, and addressing social inequities within and across countries. The task is complex, but we know what can be done, and we can build on and learn from policy and technological innovations from around the world.

—

Pierpaolo Cazzola is Director of the European Transport and Energy Research Centre at ITS-Davis.

Driving between the Bay Area and Sacramento has long been a challenge. For the past several months, it has been a construction nightmare. Heading east from Davis, drivers face seventeen continuous miles of construction zones, first for the Yolo I-80 Pavement Rehab Project , then for the on-going “Fix50” project through the core of Sacramento. Heading west, the construction zone for the Solano I-80 Managed Lanes Project  starts before Vacaville and continues through Fairfield. I dread the day that construction starts between Davis and Vacaville, not just because construction makes for a miserable driving experience but also because it means the last remaining remnants of the iconic oleander median will be gone.

What does all this construction get us – and is it worth it?

The official rationale for these roads projects is the need to modernize old freeways (especially Highway 50) and accommodate the growing flow of traffic between Sacramento and the Bay Area driven by differentials in housing costs, the post-COVID ability to work remotely, and other economic forces. Congestion “relief” is cited as a goal of the projects in both official documents and statements by public officials. To relieve congestion, Caltrans is adding “managed lanes” to freeways, usable for free by carpools and electric vehicles with clean air vehicle decals and, in some cases, by other drivers for a fee.

This rationale does not entirely square with the state’s Climate Action Plan for Transportation Infrastructure (CAPTI). One principle of this plan is “promoting projects that do not significantly increase passenger vehicle travel.” As ITS-Davis researcher Jamey Volker and I have documented, a growing number of rigorous studies provide robust evidence that adding highway capacity generally leads to an increase in vehicle miles of travel (VMT), a phenomenon known as “induced travel.” Although the evidence on the effects of managed lanes is slim, there are good reasons to believe that these lanes may generate as much of an increase in VMT as the old-fashioned “general purpose” lanes. In short: adding lanes, even if they are managed lanes, is not consistent with state goals for reducing vehicle travel.

This can be a challenging idea to accept. For the past several years, Jamey and I have participated in countless Zoom meetings in which state, regional, and local officials have argued that highway widening projects within their jurisdictions would not lead to an increase in VMT or, at most, would result in a much smaller increase than studies suggest. After Caltrans official Jeannie Ward-Waller raised questions about the Yolo I-80 Pavement Rehab Project, which appears to be laying the groundwork, literally, for an additional lane, she was fired by the agency, as reported by the Los Angeles Times. Her concerns centered around potential abuses of the environmental review process, the purpose of which is to fully vet the likely impacts of proposed projects, including impacts on VMT.

Our own work has demonstrated how induced VMT has traditionally been underestimated or even ignored in the environmental review process. Agency forecasts of the VMT impacts of proposed highway projects are not especially good at capturing the various adjustments that drivers make when a new freeway lane opens. For example, deciding to drive to a more distant store, or making a trip that previously would have been too much of a hassle. By failing to account for these adjustments, forecasts tend to understate induced VMT which means that they understate environmental impacts and overstate congestion reduction. These biases undermine good decision making. Highway projects around the country are being challenged on this basis.

All of this controversy reflects something deeper and even more difficult to address than forecasting practices: an entrenched way of thinking that perpetuates the highway-widening treadmill that has dominated transportation planning for the last century, as I discuss in my new book, Shifting Gears: Toward a New Way of Thinking About Transportation.

One idea at the core of the traditional way of thinking about transportation is that congestion is bad, that it should be reduced, and that it can be reduced. No one likes congestion, but you’d think that, after a century of failed efforts to reduce it, we would have learned better by now. Congestion did abate temporarily during the COVID-19 pandemic, but not for reasons we would ever want to repeat. Rather than focusing on reducing congestion, we might be better off focusing on ways to make congestion less relevant to our lives. One way to do this is by providing good alternatives to driving, such as high-quality transit service. Another way to do this is by bringing the places we need to go closer to where we live. Rather than pouring billions into reducing congestion for workers who can’t afford to live near their jobs, why not pour billions into providing affordable housing for them near those jobs?

Another core idea that needs rethinking is that speed is good. Speed is psychologically complicated, in that it both thrills us and terrifies us. And terrify us it should: the faster that cars go, the more likely that their occupants—and bystanders—will die when cars crash. This trade-off is well documented and widely known but, too often in practice, speed takes precedence over safety. The transportation profession defines efficiency in terms of speed, and efficiency is something to be maximized. But just how much speed do we need? Isn’t this obsession with congestion really about our expectations about how fast we should be able to get places? Is it reasonable to measure traffic “delay” with respect to traffic-less “free flow” conditions? It would be a whole lot cheaper—and safer—to reset our expectations than to widen our freeways.

The power of these ideas might explain why the cost/benefit analyses that inform decisions about highway projects also fail to account for the costs that drivers incur during the construction itself. This failure suggests that such costs are accepted as a necessary part of providing congestion relief. But as a recent Sacramento Bee story poignantly documented, the human toll of highway construction can be intolerably high. Following the publication of the article, in which I was quoted, I received a heart-breaking email from a mother who her lost her daughter last year to a crash in the Fix50 construction zone.

We owe it to ourselves to reject a way of thinking that prioritizes congestion reduction and puts speed ahead of safety. It is high time for a new way of thinking about transportation.

 

---

Susan Handy is  the Director of the National Center for Sustainable Transportation at UC Davis. Her research focuses on the relationships between transportation and land use, particularly the impact of land use on travel behavior, and on strategies for reducing automobile dependence.

Biofuels are an important tool to help decarbonize our transportation system, and their role will likely grow in coming years. New tax credits authorized under the Inflation Reduction Act are being finalized; these would offer significant incentives for the biofuel alternatives to conventional jet fuel, so-called “Sustainable Aviation Fuels” or SAF. But it’s not always clear how sustainable these fuels truly are, or whether they offer a significant GHG advantage over petroleum. The proposed SAF tax credits limit eligibility to fuels that offer at least a 50% reduction in life cycle GHGs compared to petroleum and give additional incentives to those that exceed that threshold.

Advanced fuel technologies – such as those that use inedible wastes, hydrogen, or fuels synthesized using renewable electricity – may eventually deliver very low carbon fuels, but they have yet to emerge at commercial scale. This means that current biofuel technologies that use food crops like corn or soybean oil, are likely to receive most of these tax credits for the next several years, at least. Assessing the GHG benefits of current biofuels requires a descent into the wonky world of lifecycle analysis—measuring the total emissions through farming, processing, and consumption. Much of this is fairly straightforward, by the standards of researchers and expert analysts, until one gets to the use and displacement of land.

While crop-based biofuels can reduce GHG emissions when used in place of petroleum fuels, they compete against food crops for arable land. When biofuel production increases demand for agricultural commodities like corn or vegetable oil, growers often meet this demand by expanding their planted area. Clearing land for cultivation releases much of the carbon stored in plants and soil into carbon dioxide, a greenhouse gas and the primary driver of climate change. This process of biofuel demand causing land conversion is known as Indirect Land Use Change (ILUC), and it has been a point of deep contention around biofuels over the past 20 years.

Experience has taught us that we ignore ILUC at our peril. European attempts to increase the use of biodiesel in the 2000’s overlooked the issue of emissions from land clearing. This led to widespread slashing and burning of tropical rainforest to expand palm oil plantations, much of it on sensitive high-carbon peat soil. Over 6 million hectares of tropical rainforest was lost in Indonesia and Malaysia between 2000 and 2012, demand for biofuels accounted for as much as half of this. Emissions from this conversion dramatically outweighed the benefits from additional biofuels.

Over the next few years, the key question for policy makers is: should we incentivize the consumption of more crop-based fuels, such as those made from corn or soybean oil and, if so, how much? Understanding the impacts of ILUC and getting the question right is essential to ensuring our climate policies actually reduce emissions.

All models are wrong, some models are useful

Estimates of ILUC arising from crop-based biofuel production are highly uncertain. The lowest estimates rate crop-based biofuels as up to 50% less carbon intensive than petroleum gasoline or diesel, while the highest estimates suggest that they’re several times more carbon intensive than petroleum.

Estimates of ILUC factors found in literature for a variety of fuels. Height of the gray bar represents the mean, black “x” represents the median, with uncertainty bars covering the full range of ILUC factors, and the numbers showing the number of studies contributing data points to each category. Red line shows the typical carbon intensity of petroleum. Source: Woltjer et al. (2017)

This uncertainty is due to three main factors:

1. Complexity. Accurately modeling ILUC means accurately modeling the entire global agricultural commodity system, and the decisions of millions of individuals that make it up. Reducing this complex, dynamic system down to a single, fixed number is an impossible task, but nonetheless necessary to create a stable regulatory structure.
2. Subjectivity. There’s no perfectly objective method for analyzing emissions from complex systems, like biofuel production or ILUC emissions. Modelers must set system boundaries and allocate impacts between products. Soybeans, for example, yield oil and high-protein meal, used as food for livestock or people. Assigning energy and pollution burdens to each of these is a subjective process. Two analyses of the same product can use equally valid assumptions and result in widely differing estimates of GHG impact.
3. Calibration and validation. Developing ILUC models requires calibrating against real-world data about how growers make decisions about what to grow and where. These data are often incomplete, non-public, hard to interpret, or unavailable. More importantly, historical data don’t include future climate change impacts. We know that rising temperatures and changing weather patterns will make some highly productive growing areas too hot or dry to maintain yields, and other areas will become fertile. These factors will change how and where growers choose to expand cultivation – which means today’s ILUC model predictions will necessarily be calibrated with unrepresentative data.

Due to these challenges, estimates produced by any ILUC model will be rough representations of a dynamic system, based on subjective assumptions, and calibrated against data that does not reflect the world we’re trying to make policy for. In plain language: they’re going to be wrong. The thing is, there are a lot of different ways to be wrong, some of them worse than others. The question for policy makers in this case is: What’s the right way to be wrong

How to Make Decisions When Models Are Wrong

A perfect fuel policy would provide enough support to crop-based biofuels to maximize near-term decarbonization, but not so much that demand for agricultural products soars and exacerbates ILUC-driven emissions.

In a perfect world, we would use a perfect ILUC assessment to perfectly align incentives with real-world impacts, but that’s not possible in this case, so it helps to think through what happens when we’re wrong. Overestimating ILUC’s impact means biofuels will have higher carbon intensity scores on regulatory assessments, and so fewer of them will be eligible for credits and we would expect to see less of them enter the market; overestimating ILUC would therefore lead to consuming less than the theoretically optimal amount of crop-based biofuels. Underestimating ILUC’s impact means biofuels will have lower carbon intensity scores, more would be eligible for policy support and we would expect to over-consume them compared to a theoretical ideal.

If we get biofuel policy wrong in a way that causes us to under-consume biofuels, then we miss opportunities to reduce GHG emissions. Every billion gallons of renewable diesel made from soybean oil could offer the opportunity to reduce GHG emissions by about three million tonnes of CO₂ equivalent.^[1] This is not a trivial or meaningless loss. We know we must reduce emissions very quickly in coming years in order to achieve carbon neutrality by mid-century, and soybean oil based biofuels have demonstrated commercial success already.

There are significant GHG impacts of over-consuming biofuels, as well. Each billion gallons of soybean oil based renewable diesel requires about 15 million acres of land to grow – roughly the size of West Virginia. If this land were grassland in the U.S. Midwest before being converted to cultivation, the GHG emissions from land use change alone would be over two million tonnes of carbon equivalent, and much more if it were forested or land with high-carbon soil that was converted.

These direct GHG impacts are only part of the story, however. As we dig deeper into the risks associated with over- and under-estimation, critical differences arise.

1. Timing. Carbon in land accumulates slowly but is lost quickly. Plants remove CO₂ from the air over time, storing it as solid carbon and accumulating as organic matter in the soil. Converting natural land to cultivation releases that carbon in a matter of weeks to a few years. Once it’s lost, it often takes decades to recover, if it recovers at all.
2. Diminishing returns. Higher-yielding land is likely to be converted from carbon-storing natural cover to crops first. Each additional million acres of cultivated land is likely to yield a bit less than those that were converted before it. As total demand increases, the additional land needed for every extra unit of production is expected to increase.
3. Political momentum.  It is easier for politicians and regulators to give support to businesses than to withdraw it. Removing support can also reduce the effectiveness of future climate policy; if governments signal, via policy incentives, that investors should back biofuel projects and then quickly withdraw the incentives, those same investors would be justifiably skeptical about trusting other climate policy incentives. Starting off with lower levels of support and adding mores in the future, on the other hand, avoids sending mixed signals to the market.

These issues show that the risks entailed from under- or overshooting optimal biofuel production volume are not equal. Insufficient support for crop-based biofuels could cost us the opportunity to reduce emissions, but excessive support for them could be significantly worse. Land conversion causes a loss of stored carbon that cannot be reversed on a time scale that allows us to meet mid-century GHG targets. Once over-production of biofuels occurs, any attempt to correct it could lead to stranded assets and break financial markets’ trust that policy signals are a reliable guide when making investment decisions.

We know that ILUC analysis is difficult to get right, and we cannot count on developing a model to predict the optimum level of crop-based biofuel consumption any time soon. If we can’t be sure we’re right, we should make a decision that recognizes the profoundly asymmetric risks around biofuel consumption. The risks entailed with underestimating ILUC impacts are far worse than the risks of overestimating them, so we should err on the side of overestimation.

Developing a Risk-Aware Approach to Biofuels

In practical terms, this means the arguments over which ILUC model is best are not terribly helpful since every model is going to be inaccurate. Instead, we should look at the full range of ILUC estimates we get from the many approaches of ILUC estimation that have been put forward by researchers. We’re relatively confident that the correct answer lies somewhere within this very wide range. The risk-aware approach to biofuel policy is to select a number high enough that it’s very unlikely we underestimate the real value.

“High enough to not be an underestimate” is still not precise enough to be helpful. Other landmarks can help us get closer to a risk-aware ILUC impact estimate. We know, based on extensive scientific research, that fuels from wastes and residues generally have lower ILUC impact, and that fuels from palm oil are very likely to be worse than the petroleum they try to displace. An effective ILUC assessment should align with the ample scientific evidence on these topics.

Support for additional research and modeling can help us both narrow the range of uncertainty around this issue, as well as identify what ILUC impact value within that range should be chosen to yield the best possible result. Expert-driven consultative processes, like the National Academies committee on biofuel life cycle assessment or CORSIA workgroups developing sustainability and GHG assessment methods can serve a vital role here, not only as clearinghouses for the latest research in this space but also to allow stakeholders to engage with the process. These groups can find mutually acceptable approaches to some, though certainly not all, of the subjective decisions that are an unavoidable part of LCA.

Just as there is no perfect ILUC model, there is no simple solution to alternative fuels. We know, however, that it’s going to be virtually impossible to meet critical decarbonization targets without lower-carbon liquid fuels. Avoiding the worst impacts of climate change will require risk-aware policies that succeed without perfect modeling. In the case of crop-based biofuels, it’s quite clear that the worst risks arise when we underestimate ILUC impacts. We know that our attempts to quantify ILUC risk are likely to be wrong, but we also know that there’s a less-damaging way to be wrong in this case. Policy should reflect this reality.

---

Colin Murphy is the Deputy Director of the UC Davis Policy Institute for Energy, Environment, and the Economy, and the co-Director of the Low Carbon Fuel Policy Research Initiative there. Any statements or inaccuracies herein are solely the responsibility of the author and should not be taken as representing UC Davis or the Policy Institute.

The material in this blog was originally presented as part of the EPA National Center on Environmental Economics seminar series, a recorded version is available via the UC Davis Low Carbon Fuel Policy Research Initiative website. The author would like to gratefully acknowledge the assistance of Amber Manfree, Dan Sperling, and John Schmitz in helping develop and refine this material.

 

^[1] Assuming 65 gCO2e/MJ carbon intensity (derived from approximate average CI of such fuels under California’s LCFS), and that each gallon of renewable diesel displaces 1 gallon of petroleum diesel with 91 gCO2e/MJ carbon intensity (approximate average of U.S. diesel supply).