Reducing freight emissions requires a holistic perspective.

For some time now, freight transport has been touted as the next milestone in reducing CO2 emissions. Worldwide, freight transport is responsible for 11% of greenhouse gas emissions. In California, its share is about the same at around 10.5%.

Most measures to reduce emissions in freight transport focus on reducing the carbon intensity of individual modes of transport (truck, rail, shipping, etc.). However, these efforts can be undermined if they dramatically change the cost structure of transport, shifting freight from relatively low-carbon modes of transport to higher-carbon modes of transport. But how often does “mode switching” between transport options occur? Does the choice of mode of transport respond to costs, or do goods more or less stay on their mode of transport.

This was one of the questions that served as a starting point for a recently published paper by Colorado economics professor and transportation guru Jon Hughes and me. Jon will present this work in a webinar on August 14. The research uses a fascinating dataset called the Commodity Flow Survey (CFS), which is produced jointly by the US Census Bureau and the Department of Transportation Statistics. The CFS is collected every five years and contains detailed data on millions of individual freight shipments within the US. We worked primarily with data from 2012 and 2017, which fortunately showed useful differences in fuel prices both over time and across the country.

Just examining this data is a fascinating exercise. Carbon intensity per pound shipped varies widely. High-value, lightweight goods like precision instruments have a huge carbon content per pound because so much is shipped by air. Low-value, heavy goods like coal and gravel are more likely to be shipped by rail, and so look pretty good in terms of carbon emissions per ton shipped, but less so in terms of carbon intensity per dollar of value.

There are several categories of goods that surprised me in terms of the variety of transport modes. A wide range of goods are transported by both rail and truck, which is the most likely dimension of substitution. Many vehicles, fuels, fertilizers and sand (?!?) are transported by both rail and truck. On the other hand, a significant share of the transport of precision instruments, pharmaceuticals and (?!?) live animals is transported by air. Of course, shorter distances are more likely to be transported by truck, but even taking distance into account, there seems to be quite a substitutability of transport modes for many goods.

Jon and I were interested in mode substitution because many climate policies target specific modes individually. Our original work was motivated by Obama-era regulations that increased the required fuel efficiency of many trucks. California has mandates to electrify both trucks and rail by the middle of the next decade. All of these policies could change the cost structure of transportation, essentially making modes more expensive due to higher shipping costs, but lowering the marginal cost of transportation per ton-mile due to economies of scale.

This is related to the more general phenomenon of “rebound”, where improved energy efficiency (e.g. less fuel consumption) leads to more frequent use of the product (more kilometers). The topic has been discussed both empirically and theoretically, as rebound is not necessarily a bad thing (more comfort!, nicer holidays!), depending on what interests you. However, from an emissions perspective, rebound can recoup some of the savings achieved through improved efficiency through more intensive use.

We were curious about another dimension of rebound, which we called “cross-rebound” to describe the prospect that improving the efficiency of one type of freight would divert freight from other modes of transportation. Cross-rebound is likely a subset of general rebound, since a measure of, say, increased truck miles includes both new miles and miles saved from other modes of transportation.

The CFS gave us the opportunity to isolate this mode choice effect (or cross-rebound effect). We examined how the choice of mode of transport for certain goods changes with fuel prices. Intuitively, we would expect these effects to be very different for different goods. The transport of textiles, paper, machinery and chemical products was very sensitive to the relative transport costs per mile of the different modes of transport. In contrast, when shipping heavier goods such as coal, unprocessed grains and vehicles, the choice of mode of transport changed little with the relative costs of the different modes of transport.

To study the impact of a regulation requiring improved fuel efficiency of trucks, we tried the following maneuver. First, we calculate emissions assuming that all transports remain the same mode of transport, but that trucks become 5% more efficient. Then, we calculate emissions assuming the same efficiency improvement, but allowing transports to switch modes if lower fuel costs make it desirable. Since most of this substitution is from rail – which is comparatively less carbon-intensive – to more carbon-intensive trucks, these switches offset some of the emissions reduction.

Again, the effect is very commodity specific. The rebound effect for more reactive goods like paper, sand, and basic chemicals is quite large. Our estimates suggest that cheaper truck miles will drive a lot of traffic away from rail for these goods. On the plus side, some precision instruments and live animals (?!?) will shift from extremely carbon-intensive air freight to trucks, increasing base profits. Overall, however, our estimates suggest that about 20% of the emissions reduction could be recouped by a five percent improvement in truck fuel efficiency from this modal shift from other modes to trucks.

This is not to say that all efficiency efforts are bad or that all rebound effects worsen the emissions balance. These types of effects depend on how specific measures affect the cost per ton of transport in a particular mode and on the overall pattern of cross-elasticities.

Full electrification of freight rail is an interesting example. It’s possible that the marginal cost of moving a train per mile is lower if it’s powered by electricity (as long as you don’t buy it in California). However, if the increased capital cost of rail equipment means there are fewer trains overall, that shortage could drive up transportation prices per ton-mile.

Our main conclusion is that these impacts are significant and that measures targeting specific modes or technologies risk unintended consequences. This is a good time to add the economists’ point, required by the Treaty, that technology (and mode) neutral measures that send a consistent signal to all sectors, such as an emissions trading scheme or a carbon tax, have a better chance of minimising inefficient cross-rebound effects.

—

Follow us on Bluesky and LinkedIn and subscribe to our email list to stay updated on future content and announcements.

Recommended citation: Bushnell, James. “Will the CO2 reduction continue like this?” Energy Institute Blog, August 122024, https://energyathaas.wordpress.com/2024/08/12/will-carbon-reductions-keep-on-truckin/

—

By default, comments are displayed anonymously, but if you are happy with this, we recommend that you sign your comments.