Better Freight Policy

Over the past few months, American policymakers and transportation scholars have been debating a potentially sizable infrastructure package. In urbanist circles, most discussion has centered around the relative quantities of transit and highway funding, ways the transit funding might be spent, and potential associations between infrastructure monies and land use reform. These topics are unequivocally important, and if anything merit more mainstream visibility than they have received. Lost in this milieu, however, is the importance of improving America’s approach to freight.


For the last hundred years, America’s approach to goods movement has favored trucks. Able to move more quickly and flexibly than rail or barge, and aided in their growth by federally and state-funded road improvements, a regulatory structure that gave truckers significantly more flexibility to pursue traffic than railroads, a national interest in decentralizing industry, and a societal disinterest in addressing the externalities of road use, truckers emerged from the twentieth century having captured a plurality of the domestic freight market. To some, this shift might look impressively small. Compared to the fate of American passenger rail, American freight rail has remained extremely strong. In international context, America continues to look like a success story, as the strong fundamentals of the American rail network staved off the sorts of mode share collapses experienced in Europe. Yet this image is deceiving: by most measures, freight’s pendulum has swing too far. This ‘truckers’ century’ — and the continued upwards creep of truck mode share — has left us with a freight system that is increasingly an agent of crisis.

Trucks’ damage is most immediately apparent in the fight against planet-warming and community-harming emissions. The interface between rubber tires on asphalt may allow road vehicles unbelievably versatility, but it also makes them inefficient: compared to trains, trucks emit ten times more CO2, five times more NOx, two times more PM10 and 1.3 times more PM2.5 per ton-mile. Put differently, trucks haul 39 percent of American freight and trains about 33 percent, yet trucks generate nearly eleven times more carbon emissions than rail, or about a quarter of all (passenger and freight) transportation-related emissions in the United States. As supply chains have lengthened and sped up, this impact has grown dramatically: since 1990, carbon emissions from trucking have been the fastest growing of any major type of transportation. Fleet electrification may begin to reverse this trend in coming years, but carbon is not trucking’s only environmental cost: tire- or brake-driven particulate matter emissions will remain a major source of air pollution after electrification. Especially in urban areas, where congestion and racialized geographies of poverty and transportation mean that emission impacts disproportionately hurt low-income communities of color, trucks participate in a transportation system killing our planet and aggravating disparities in our welfare.

Preventing catastrophic climate change and health crises are inarguably the most immediately important challenge for freight policy, but it is hardly its only salient problem. Because road damage scales approximately with the fourth power of axle load, trucks and other heavy vehicles are responsible for most use-related wear on America’s roads. As anyone who has ever driven down New York’s BQE can attest, this can wreak havoc anywhere that trucks are present and maintenance dollars are not, but the infrastructure impacts of trucking are today most acutely felt in rural America. In counties where shrinking populations must support the ever-increasing costs of heavy trucks (ones which will only grow with electrification), poor road quality has begun to severely impair the basic functioning of transportation systems. As rail service grows ever harder to access and the relative cost of trucking remains low, rural residents and industries increasingly exist in a world of infrastructure decay and failure, with alternatives and mitigations not much more than a speck on the horizon.

The more particularly urban of trucks’ impacts on the road network are their impacts on road safety. Thanks to their size and blind spots, and the fact that they do not have (largely) dedicated infrastructure like trains or barges, they pose a significant danger to the people around them. Though their disproportionately large share of road injuries is visible across geographies and types of accident, it is especially problematic in cities where trucks mix with relatively high levels of bicycle and pedestrian activity. In New York, trucks are responsible for 32 percent of cyclist and 13 percent of pedestrian fatalities despite being only about 10 percent of vehicle traffic. Risks associated with trucks in cities seem likely to grow dramatically in the coming years, as e-commerce growth puts more and more last-mile goods movement on trucks and brings more long haul truck trips to urban warehouses. Now more than ever, it is thus critically important that we find safer ways of moving freight.

Finally, trucks are deeply implicated in a geographic system that has played a leading role in producing concentrated and racialized poverty in American cities. As I discussed previously, trucks were key to the decentralization of industry in the twentieth century, allowing business to flee cities for greener pastures. In combination with the structural discrimination written into American housing and transportation policy, subsidized and truck-enabled decentralization amounted to a massive denial of opportunity. Rapid urban deindustrialization deprived cities and their disproportionately minority residents of well-paying industrial work at the peak of American Fordism, helping abet societal inequalities that persist to this day. Though rising demand for package delivery has caused some resurgence in urban warehousing, industry generally continues to sprawl, further cementing trucking’s dominance.

These crises of freight have no simple solution. Goods movement is complex, rooted in physical path dependencies, and inseparable from broader societal crises. To the extent that there does exist hope for change, it is (as other bloggers, advocacy groups and government climate strategy documents have argued before me) through policy action to shift freight off of our roads, and onto trains and boats. Because the American rail network is the only real alternative to trucking for most cargo flows, this post will focus specifically on shifting goods to rail — but rest assured that ships and barges have an incredibly important role to play in improving goods movement. As is commensurate with the complexity of this problem, mode shift strategies will have to be multifaceted. Freight’s issues are rooted in the convergence of infrastructure, land use, policy, and shippers’ choices; remedying its problems will require changing incentives, making investments, and reforming America’s industrial geography.


Neoclassical economics teaches one to address imbalances in systems like freight movement by internalizing externalities, or making things’ private costs reflect their total costs to society. Thanks to the political difficulty of increasing the price of popular things like gas — to say nothing of the fact that social cost calculations are often riven with epistemic uncertainty — we will never live in a world with perfect price incentives. However, because freight does not vote (ever heard of the tire tax? Yeah, I didn’t think so.) obtaining better outcomes through improved road pricing may be a politically feasible means of offering shippers better mode choice incentives. The potential impact of pricing is rather significant. A CBO working paper sought to add up all the above-discussed emissions, road maintenance, congestion and safety impacts related to trucking, and weigh them against truckers’ tax payments. Even with a rather conservative cost of carbon, the study found that truckers underpay for the privilege of driving on American roads by about 15-30% — a gap significant enough to have major ramifications for mode split.

Because of the time-, place- and weight-related variability of trucking’s impacts, the most economically efficient pricing scheme to remedy this gap would rely on tracking and load sensors in all of the nation’s trucks. Complex as this sounds, the Swiss managed to implement such a system in the early 2000s. Under their law, all goods vehicles over 3.5 tons must have a tracking unit, allowing precise and constant tracking. The fee seems to have been a major success: the Swiss transport ministry credits it with helping the country’s efforts to control the growth of truck traffic on the country’s transalpine highways, and indeed trucks’ share of freight traffic has declined since the phase-in of the fee in the early 2000s.

In a future world dominated by electric vehicles, some such system of VMT taxation may be desirable lest we lose price disincentives for road use. However, as the battles over the infrastructure package have made painfully clear, the politics and administrative complexity of such measures pose rather significant challenges to more immediate reform — especially in the United States’ low-trust environment. A tracker-based fee levied only on trucks as in Switzerland may be moderately more salable, but still would have high administrative costs in a country with about 12.5 million heavy commercial vehicles (past efforts to rely on self-reported mileage have led to significant rates of tax evasion). In its stead, we might consider incrementally expanding tried-and-true methods of pricing road access like tolls. The legislation authorizing the Interstate Highway network was written in an era when policymakers expected gas taxes to pay for road upkeep, so to prevent governments from overcharging users, they banned tolls. In the interceding decades, that provision has been modified slightly to allow tolling on new capacity (hence the proliferation of HOT lanes), but with gas tax revenues set to shrink further, now seems a potentially ripe occasion to repeal the provision in its entirety. An expansion of tolling to cover more of the American highway network could play an important role in our efforts to reduce truck use.


What you do with a pool of revenues is as important as how you obtain it. Historically, road user fees have functioned as something of a slush fund for unfettered highway expansion with little regards to social impacts both within (congestion) and without (emissions, safety) the highway network. Given the high environmental costs of increasing road vehicle use and the urgency of our present climate crisis, we should be working to stop expansions as soon as practicably possible. On the flip side, the century-long emphasis in American policy on providing (relatively) abundant capital to improve roads — and, for the most part, only roads — has contributed to conditions of relative capital starvation on the rails. For freight policy to be successful, this must change: radically expanding the amount of government spending on freight rail will be central to better goods movement outcomes.

Railroads are unique among the major modes of transportation in that they overwhelmingly own and maintain their own infrastructure. Government provides roads for trucks and cars, locks (and often ports) for barges, and airports and air traffic control for airplanes — but plays no significant role in providing the social good that is freight rail transportation. Railroads do a good job maintaining their sprawling networks, and their fortunes have certainly improved since the depths of the 1970s, but the cold reality of the industry remains one of long-term capital starvation. Only in the late 2000s did railroads begin to earn returns which exceeded the industry’s cost of capital — an impressive feat, but one whose achievement seems to have involved a focus on price increases and demarketing from insufficiently profitable business segments. With rail’s mode share now in decline, financial health seems to have come at the cost of traffic growth: railroads do not wish to face the capital costs of volume gains, especially given the immense geographic shifts in capacity that will necessarily follow a pivot away from a coal-and-bulk-goods-oriented network.

Improved pricing would indubitably induce more railroad capital spending, but achieving mode share at the scale and speed we need while ensuring that plans advance other equity and environmental goals will require government investment. On a most basic level, progress on pricing is by no means guaranteed, and it is even more unlikely that pricing schemes would come close to accurately approximating the full costs of trucking. But even if such a perfect scheme were implemented, there still would exist important reasons for governments to spend on rail. Railroads have historically struggled to resolve collective action problems, aggressively pruned feeder networks, and protested most asks for cooperation with passenger operators. Moreover, if the last decade of railroad history is any guide, institutional path dependencies may lead to more price- rather than volume-centric strategies to realize railroads’ market power from increased truck tolls. Perhaps most importantly, private rail investments made in conditions of current-day prosperity exist in a fundamentally unbalanced infrastructural system; railroads’ century of disinvestment and neglect has made for an underdeveloped base network. The task before policymakers thus is one of working with the railroads to facilitate catch-up rail network development, lest rail’s mode share continue to slip.

A good sanity test of this dual pricing-investment proposal is that it is the approach increasingly adopted in nations with stronger transportation-climate policies. Two thirds of the revenues from the Swiss heavy vehicle fee is spent in support of freight rail, including the famous New Rail Link through the Alps program. German truck tolls flow back to roads, but the German government has stepped up public investments in its freight rail network recently as a part of a plan explicitly targeted to increase rail mode share. If anything, American projects have more potential to shift goods flows than their European counterparts: American freight railroads are significantly more efficient than those across the pond, not being limited by extremely outdated coupler technology or constraining infrastructure. This increases the potential social benefits of investment, and makes current inaction all the more dangerous.

Chicago’s CREATE program provides a good, real-world example of what a freight improvement program might do. Unveiled in 2003 as a solution to that region’s century-long rail congestion problem, the program was a joint effort by governments and the region’s passenger and freight railroads to create an investment plan which would help improve operations in the metropolis. Its contents are rather unspectacular — new flyovers, improved junction configurations, modernized signaling and the like — but seem set to be among the highest value major infrastructure investments made in recent years: 2013 projections estimated that the program as a whole would yield a benefit/cost ratio of 6.3. Though most regions don’t share Chicago’s particular set of problems, CREATE’s emphasis on solving low level operations issues is instructive. Especially in the South and West, freight rail infrastructure remains underdeveloped for current needs; many terminal areas (like Houston) are virtual warrens of flat crossings and single track. This philosophy’s relevance extends beyond congested urban centers, too. For example, much of the rural American rail feeder network is struggling [link is a PDF download] to upgrade track infrastructure to handle increasing railcar wrights; a national effort to replace aging bridges, rail and culverts would likely yield significant dividends to shippers and road departments at relatively little cost. The upshot here: with the rail system so desperately in need of capital, there will be few silver bullets for freight. Rather than trying to find our own Gotthard Base Tunnel, American governments are likely best served preparing for an onslaught of the truly mundane.

Equally important as the basic need to radically increase our investment in sustainable freight transportation is the existence of a rigorous framework for planning. Switzerland and Germany both have set national level freight policy goals which guide their network investments. A revised version of the National Strategic Freight Plan with better-defined objectives could serve as an equivalent document for the United States, providing a government-wide framework in which actions might be evaluated. However, because state or regional bodies mediate most infrastructure funding, organizational action on those levels is in some ways more important. Mandating that metropolitan planning organizations (MPOs) engage with freight railroads and prioritize mode shift in their transportation network investment planning would help advance freight goals immensely, and allow closer coordination with regional transit and land use visions — something critical to the successful, aggressive and simultaneous expansion of freight and passenger rail in the US.

Finally, freight policy should encourage improved rail operations to accompany infrastructure spending. The past half-century of American railroading has been defined by increasing sectoral specialization in long haul (>750 mi) bulk freight movements. This focus has increased profits and productivity, but it also means that railroads are increasingly weak competitors in the short(er) haul, higher value parts of the freight universe which hold most traffic. Railroads have even found ways to transform small shipments into big ones. It is unrealistic to expect railroads to become competitive in all movements or markets, but government should encourage railroads to expand their offerings. Notably, there does exist precedent for this sort of involvement in rail. During the industry’s worst years in the ’70s and ’80s, the Federal Railroad Administration provided funding for a short-haul, high-speed, high-frequency intermodal experiment, which became the Milwaukee Road’s highly successful fleet of Sprint trains connecting Chicago and the Twin Cities. Reviving this sort of involvement through loan or grant programs could de-risk new freight services — whether that be oft-discussed short haul intermodal services, the critically important task of improving loose carload offerings, efforts to better coordinate with other supply chain actors, or otherwise — helping railroads push beyond their institutional biases towards slow and long freight, and thus improving the social return on rail investments.

Land Use

The final piece of the freight puzzle is industrial geography. Much as transit ridership increases with population density and employment centralization, freight rail use increases as shippers cluster near tracks and terminals. This is true for two reasons. First, the rail network is neither comprehensive nor porous: a (for example) tank car full of syrup can travel on the 140,000 miles of American rails, and no further; an industry wishing to receive said tank car must locate along a rail line to get it. Secondly, railroads fundamentally are about scale economies: they assemble multiple shipments into one train to move freight. As industry decentralizes, it becomes less likely to be located along rail lines, and harder to serve with rail: trains’ scale economies weaken when traffic is less concentrated, and when feeder movements to and from terminals consume more of total travel. To ground these abstractions, one need look no further than documents detailing the decline of freight railroads in the 1970s: they list industrial decentralization as among the key causes of the sector’s decline. Consequently, centralizing industry near rail lines — and especially near rail terminals — will be key to successful mode shift policy.

Fixing this tendency towards sprawl will be no small task. On top of the general difficulties of passing planning legislation of any sort in this country, local control of land use decisions make any sort of coherent regional — let alone national — approach all the more elusive. As Sandy Johnston noted in his (excellent) post on industrial sprawl, plant location has also become a site of intense intermunicipal competition. In pursuit of improved economies and tax rolls, localities often offer significant incentive packages to companies locating within their city limits, further complicating the politics of industrial sprawl.

The way forwards lies (again) in multifaceted action. As Sandy suggests, we need tax reform, better brownfields remediation programs, and rail-proximate plant location incentives. Additionally, expanding programs which provide grants to shippers and industrial developers who wish to connect to the rail network might further encourage location near rail lines. On the flip side, easing restrictions in current zoning laws to allow higher density industrial development (which is back in vogue these days) also may help maximize the potential of legacy urban industrial areas, which often are extremely well located for rail. This is not to say that we should begin building steel mills in Bushwick or try to bring meatpacking back to Chicago’s South Side; compromising environmental, economic and mobility justice goals for a better freight mode split would be terrible policy — but that, wherever reasonable, we should work to maximize the amount of supply chain activity along rail corridors.

What a rail-centric industrial future might look like is already becoming visible. One of the largest American railroads (BNSF) has recently taken to developing “logistics parks” around some of its intermodal terminals, siting warehouses to minimize rail access costs and create something of a captive customer base for BNSF’s trains. Touted both for their environmental and cost savings (less mileage between the loading dock and the railhead economizes both on money and carbon), this model provides some amount of rail-centrism in development. Though this may offer a much-improved model for locating space-intensive industries, it is only a partial solution to the broader ills of poor freight policy, because these logistics parks are generally built at the leading edge of metropolitan sprawl. Exurban locations offer unbroken spaces for construction, but they also make for long truck trips to (sub)urban consumers, and poor accessibility from low-income deindustrialized communities in urban cores.

Though certainly an ‘easier’ reach from current land use conditions, we should not settle for some variety of controlled sprawl. Adopting a more Swiss freight land use vision could provide even better freight policy outcomes. The edges (and often the interiors, too) of Swiss cities are littered with higher density industrial estates, where freight rail directly serves warehouses and other businesses. This variety of development — reminiscent of old American cities — builds on the strength and reliability of Swiss direct rail services to create noticeably truck-light landscapes. This, in turn, facilitates densified development: fewer trucks means less parking and fewer overlarge roadways. Though these landscapes often are suburban, their density makes them more spatially efficient (and amenable to transit service), presenting a potential model for a more ambitious imagination of the industrial future. The US has some significant hurdles to surmount before these sorts of developments can become mainstream (including chronic issues with our carload freight services), but a handful of congested cities are already seeing renewed interest in multistory industrial projects. Though these projects tend to be truck-oriented, there may be hope, after all.


It is easy for activists to think that their specialty holds the key to some imagined future, so I want to be clear: even the best possible freight policy cannot alone stop climate change, prevent highway-related health problems, improve traffic safety, or remedy structural economic injustices. Much as better freight policy’s components are mutually interdependent, goods movement efforts themselves will amount to little outside a larger program for climate, infrastructure and justice. Nevertheless, freight improvements hold the potential to have significant impacts. Nobody has ever rigorously analyzed the potential implications of a long-term, comprehensive freight program for the US like is proposed here, but roughly adding the estimated shifts from a few studies which look at pricing and investment options suggests that better policy could cause (roughly, and likely conservatively) 5-8 percent of all freight to shift modes — a figure which might sound small, but could potentially affect a 10-20 percent reduction in freight transportation emissions on a relatively reasonable time scale, and would have lasting benefits for safety, road maintenance, urban geography and community health far beyond that. Freight thus might be the sector in which we are best equipped to begin the overdue work of radical transportation reform: with its impacts increasing quickly and its sprawled geography hardening by the day, better goods movement policy is urgently needed.

2 thoughts on “Better Freight Policy

  1. Would the electrification of the US’ freight network be feasible? And would such a project be cost effective from this same framework you bring up here?


    1. Expensive, but certainly feasible and definitely very important to decarbonization and network reliability goals. Many studies have shown that it makes sense on the dense main lines, but I think uncertainty about diesel prices + property tax considerations + the upfront capital costs have deterred RRs from going for it.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s