At a Glance
- Remedies to address road congestion are often predicated on the presumption that road users do not cover road infrastructure costs.
- This report provides estimates of the extent to which motorists in Ontario cover infrastructure costs.
- Light-duty vehicle users cover a significant portion of road infrastructure costs. In addition, cost recovery in the Greater Toronto and Hamilton Area is likely significantly higher than at the provincial level.
Most people agree that road congestion is a problem, but there is considerably less of a consensus on what to do about it. Should we build more transit or road infrastructure? Should we introduce tolls or increase fuel taxes? Should we be more prescriptive in terms of restricting the growth of residential and commercial areas?
Economists seem to agree that we have the structure of road prices wrong, and that some form of road tolls, which vary by time and place, are part of the answer. In this case, do we simply restructure the prices in order to collect the same amount of revenue? Or do we restructure prices and, as well, collect more or less revenue? If more, how much more do we need? And what do we do with that revenue?
There are no simple and straightforward answers. But virtually all of these questions are related to a single presumption—that road users are heavily subsidized. And, as a result, road users don’t consider the full costs of their behaviour in their individual decision-making. To begin answering these questions, it is necessary to first quantify the extent to which road users are subsidized. This is the key focus of this report, using the Province of Ontario as the target of our analysis. A subsequent report will elaborate on the policy implications.
Infrastructure Cost Recovery of Light-Duty Vehicle Users in Ontario
For this report, the Conference Board has produced new comprehensive estimates of the costs and revenues of the Ontario road network. The estimates are based on the most current data and methods that reflect best practices of government agencies and independent researchers worldwide. Determining the degree to which road users recover their infrastructure and related costs is no straightforward task. This is due to data limitations and the numerous government departments and levels of government involved in providing services to road users. To determine total road infrastructure and related costs, we took three broad approaches:
- A direct, expenditure-based (pay as you go) approach.
- An annualized capital expenditure approach (where depreciation and interest on the net current value of the capital stock is added to operating expenses).
- A road inventory (bottom-up) approach.
The first two approaches relied primarily on public accounts data. The third approach relied on geographic information system (GIS) data and engineering estimates of the unit cost of various road and related infrastructure costs.
Revenues that were raised from taxes and fees unique to road users—such as excise taxes and vehicle registration fees—were used for the basis of determining road user revenues. Non-unique taxes—such as the HST charged on the purchase of the fuel or on the excise tax—were not included as road user revenues.
Furthermore, to limit the calculation to light-duty vehicle users or motorists (drivers of light-duty passenger vehicles [LDVs] such as cars, minivans, and SUVs) we required an assessment of the extent to which these vehicles impose costs on the road network, relative to heavy vehicles.
Allocating Costs and Revenues Among Users
Some costs, such as clearing the snow and debris from our roads, do not vary by vehicle type. In that case, we should allocate routine maintenance costs on the basis of the vehicle-kilometres travelled (VKT) per vehicle. Other costs, such as the creation and maintenance of the rights-of-way, do vary in some proportion with the size of the vehicle. Therefore, we should allocate costs based on passenger car equivalent units (PCU). Accordingly, a heavy truck that is approximately three times the size of the average passenger vehicle would count as three PCUs.
Some costs vary according to the gross weight of the vehicle. For example, bridge structures that are constructed to accommodate heavy trucks must be built to a higher standard. And, the life of the structure varies according to the weight of the vehicles that it carries over time. Vehicles also cause direct wear on pavement, which generates maintenance and eventual replacement costs as a result. The extent to which a given vehicle causes wear on the pavement is directly related to axle weight and not the gross weight of the vehicle. For costs associated with pavement wear, we should allocate on the basis of equivalent single axle loads (ESAL).
Allocating Revenues Among Users
A large portion of the revenues generated from road users comes in the form of fuel excise taxes. While we do not have a direct measure of how much of the fuel tax revenue is collected from motorists, we do have data on gasoline and diesel fuel consumed for on-road use. As well, we have estimates of the proportion of these fuels that were consumed by light-duty vehicles. This allows for an estimate of fuel tax revenues generated from motorists. Detailed data on other sources of revenues—such as licensing, registration, and traffic infractions—are more difficult to attain. Therefore, we have to rely on more approximate allocations, such as by VKT or PCU.
Summing-Up How Much Users Pay
Taking all of these factors from above into consideration, we estimated the Ontario motorist cost recovery for the three years ending in 2010—the most recent year for which the necessary cost data are available. Chart 1 summarizes the results.
Cost recovery is estimated to be highest when using the expenditure approach (over 80 per cent in 2010). In all three of the infrastructure and related cost approaches we took, the cost recovery has declined over the three-year period. This is due to the unusual large increases in spending—including the various recent government infrastructure and stimulus programs. Note that the above estimate does not allow for any allocation of costs to non-users. Moreover, the results mask the issue of the imbalance of revenues and expenditures by level of government. The federal government collects a significant portion of the revenues but owns and maintains a relatively small portion of the road network, whereas local governments find themselves in the opposite situation.
In the Greater Toronto and Hamilton area, the road infrastructure cost recovery is significantly higher at the urban level and is over 100 per cent.
These results provide an indication of cost recovery at the provincial level. But, for policy purposes, it is more instructive to understand cost recovery within urban areas and, indeed, even for specific links and nodes on the road network. This is virtually impossible to do using publically available data. However, we did use GIS and vehicle registration data to help estimate the road network costs in the Greater Toronto and Hamilton Area (GTHA), which is the largest urban area in the province. This allowed for an estimate using only the road inventory approach, which produced the highest cost estimate at the provincial level. Chart 2 summarizes the result for all road users in the GTHA (not just motorists).
These results are subject to greater uncertainty than the provincial estimate. However, the magnitude of the difference is such that we have confidence that the road infrastructure cost recovery is significantly higher at the urban level and is over 100 per cent. Cost recovery is greater than for the whole province, essentially due to greater traffic density and higher fuel consumption, which raise revenues per kilometre of road relative to costs per kilometre.
There are other costs associated with driving, such as the cost of owning and maintaining a vehicle, accident costs, congestion costs, and environmental costs. Vehicle ownership and maintenance costs are the most significant of these (approximately $50 to $60 billion at the provincial level). For policy-making purposes, we are less concerned with these costs because they are absorbed directly by users. But to put them into context, they are approximately $0.45 per VKT, whereas infrastructure costs are $0.06 to $0.07 per VKT. If we were to look at cost recovery in terms of recovery of infrastructure and vehicle costs (which may be more useful when comparing the cost recovery ratio to most other modes of transportation), the cost recovery would always tend toward 100 per cent as a result.
The other costs are more difficult to quantify, but we can rely on previous research to place these costs into the context of the infrastructure costs. Table 1 offers rough estimations of these costs on a per VKT basis in order to provide an indication of how large they might be relative to vehicle and infrastructure costs.
Caution must be taken with social cost estimates as they vary considerably, and small changes in assumptions (such as the unit value of time for congestion cost) can alter the estimates radically. And while it may seem counterintuitive given the magnitude of the costs, for policy purposes (at least for infrastructure policy) the accident costs are less of a concern. The reason is that they are, for the most part, met by individual users. In other words, we do not need to “charge” users for these costs as we might want to do for infrastructure and other social costs. Emissions costs, on the other hand, are pure externalities. They are generated by the group of road users and absorbed by non-users. The rough estimates above show criteria air contaminant (CAC) emissions costs to be above GHG emissions costs. But it is worth noting that CAC emissions from light-duty vehicles have been on a steady decline in Canada and are expected to continue their steady decline in the future. For example, particulate matter (PM10 and PM2.5) and volatile organic compound (VOC) emissions from LDVs in Canada were approximately half of their 2002 levels in 2011, while sulfur oxide (SOx) emissions have been declining even more quickly.
Congestion costs are in a category of their own. They may be external to the individual user but they are, for the most part, internal to the group of road users. As such, while there is an important policy rationale to minimize these costs, congestion costs should not be added to the other costs in order to determine “total” costs for the purpose of a cost recovery calculation. Moreover, the concept of an “average” congestion cost has little meaning. Congestion costs are, by definition, marginal and cannot be averaged across time and space.
Motorists in Ontario meet at least a large portion of the costs that they impose on the road infrastructure—and in major urban areas probably much more than those costs. If we look at the total cost of driving, including vehicle costs, cost recovery will tend to be closer to 100 per cent.
Our calculations can be improved upon if better data were collected and made available to the public. This is no small task given the various governments involved in the provision of road infrastructure. But data gathering should be the focus of more effort given the implications for infrastructure policy. Nevertheless, these results are a useful first step toward answering the questions about the use of user charges for efficiency purposes, revenue generation, and revenue allocation. Because all of these purposes often call for similar remedies, such as new taxes and fees, policy-makers should be explicit in the foundation and intended purpose of these remedies. In a subsequent report, we will further interpret these results and place them into the context of the broad principles that ought to govern our infrastructure policy.