Review of the Big Move: Understanding Benefits Case Analyses

Metrolinx introduced The Big Move alongside a set of guiding principles that included investing "where it matters most," prudent financial management, accountability, and risk management. Sound, sensible principles, all of them.

Metrolinx said it would subject each scheme to rigorous review, to be presented in "Benefits Case Analyses" or "BCAs." The name is similar to "benefit-cost analysis," a standard economic approach to evaluating the merits of a proposed projects. A "Benefits Case Analysis" is a report, not a methodology, but a BCA normally includes a benefit-cost analysis along with additional information to understand the case for a particular project.

In his critique of my Review of the Big Move, Steve Munro notes a number of problems with Metrolinx's BCAs, but his comments suggest some confusion over how the analysis is done and why.

In a benefit-cost analysis for transit, the first step is to estimate the costs of construction, rolling stock, operation, maintenance, and labour for the scheme (be it subway, LRT, or BRT line, or some other type of upgrading or investment). Costs may also include estimates of the effects of disruption caused by construction or increased congestion from the loss of a lane of road that has been converted to transit-only use.

These costs can then be compared to the benefits of the project in terms of revenues from additional fares, reductions in road congestion (by getting more people to use transit) or in crowding elsewhere on the transit system (by diverting existing riders to a new line), and time saved by passengers (because their time has a value). Benefit-cost analysis normally considers direct financial costs, but also imputed or implied values of things like passenger time, road congestion relief, or service disruption.

Besides purely transport benefits, sometimes there is consideration of "wider economic benefits." These might include development potential along the route, which contributes to property taxes and employment. Some analyses also estimate environmental and social benefits as well as economic benefits, such as reduced emissions from automobile use or improved access to employment and services for people who use the line.

While there is little doubt these benefits exist, they are hard to estimate. For example, development may or may not occur, and even if it does, it might have occurred anyway, or at another location. Steve Munro is skeptical about such benefits, and there is a valid argument to be made that raising taxes to pay for such benefits may not be fiscally prudent.

One important principle of benefit-cost analysis is that wherever possible, impacts should be converted into a financial value, even if money does not actually trade hands. This means estimating the dollar value of people's time, for example, or the value of reduced greenhouse gas emissions. And to allow for comparisons, all current and future costs are discounted to a "Present Value," the amount of money today that would be equal to all the benefits in the future.

Another principle is that there should be no "double counting." For example, the fares people pay to travel on a new transit line are treated as an estimate of the value to those passengers of the service. So if they pay higher fares to save time, one should not also count the value of the time they save (because the value of the time saving is already captured through the fares they pay).

Clearly, it is easier to estimate costs, because budget items such as the price of a subway train or the value of a bus driver's time can be estimated in advance with some accuracy. Benefits are harder to determine. For example, how does one determine the number of drivers who will make the switch to transit from automobile use? The answer to this question depends on the route and the density of housing and employment around each stop, the speed of the vehicles, the time required for a typical journey, and many other variables. Nevertheless, there are computer models to support such calculations, calibrated using data from experience of other transit projects.

In the end, a benefit-cost analysis can provide two crucial numbers for each transit project: (1) the "net benefits" of a project, essentially the total benefits minus the total costs and (2) the ratio of costs to benefits. These two numbers make it possible to determine which schemes offer the best value for money.

Metrolinx commissioned Benefits Case Analyses (BCA) for most of the major projects that make up the Big Move, but not all of them. It seems BCAs were never prepared for the Union Pearson Express and the subway line to Vaughan. Or perhaps they were prepared but not released to the public. The BCA for the Eglinton Crosstown line was released only after construction had begun on the line. And although Metrolinx released a "Preliminary Benefits Case Analysis" for the Downtown Relief Line, it contains no estimate of benefits and so is a BCA in name only.

Moreover, many of the Big Move proposals have been modified, but Metrolinx has been slow to publish updated BCAs. For example, the Metrolinx website still shows a BCA for a single Sheppard-Finch LRT line, but no BCA for the two separate schemes that are currently being planned.

The approaches used in the Benefits Case Analyses vary, as they were not all done by the same consultants, but several have a common format. They include a benefit-cost ratio, which compares transport benefits (usually travel time savings, automobile cost savings, accident reduction) to capital and operating costs. There are also qualitative criteria (not shown as a dollar value but estimated on a scale of one to three) for environmental benefits, social and community benefits, and economic development benefits. Cost per new rider is not calculated, and in most cases the number of new riders is not disclosed - only the total number of expected riders. One needs to make an informed guess as to how many of these are new riders, and how many would otherwise ride transit anyway.

Normally, BCAs are prepared to inform policy decisions as to which of a range of alternatives to implement, and which schemes should be given priority. While some of the BCAs do evaluate alternative schemes (for example the Mississauga Dundas Street BCA considers Bus Rapid Transit and LRT alternatives, and different lengths of scheme), most consider a restricted range of options. The Eglinton Crosstown BCA considered wholly underground schemes and schemes connecting directly into the Scarborough RT. Unfortunately, the conclusions seem not to have been understood, as the BCA shows clearly that a fully grade-separated scheme would attract many more new transit riders and have higher net benefits.

In my Review of Metrolinx's Big Move, I evaluate each proposed project using the same criteria, and present the results so that non-experts can understand their relative merits. I have estimated net benefits and the benefit-cost ratio for each project as well as the net cost per new rider for each project. This shows how much public money each scheme will cost, net of incremental fare revenues, for each new transit rider.

Steve Munro feels that evaluating the projects line by line overlooks network benefits - that is, the way in which the whole can be greater than the sum of its parts. This criticism applies to Metrolinx's own approach, in which BCAs were produced independently by various consultants for each element in the Big Move. However, there is a limit to the value of lumping together an unprofitable line with a more profitable one. If Metrolinx's own BCAs show that LRT lines will cost $10 billion to construct and yield only $5 billion in benefits, no amount of "network effects" would redeem them.

Fortunately, there are many schemes with good benefit-cost ratios, including upgrading the GO system into "Regional Express Rail." The Review also shows how the Eglinton Crosstown and Scarborough schemes might be altered to improve the ratio of costs to benefits.

Steve Munro also implies that Metrolinx's BCAs and the analysis in the Review consider only regional, and not local benefits. In fact, the analysis includes short- and long-distance trips. At the same time, however, not all riders represent equal value. It is worth more in investment to attract new riders onto transit if that investment reduces the number of car journeys made on crowded expressways; it is worth less to attract a new rider onto transit if that person would otherwise make a short trip by car on local streets (and not worth anything at all if he or she would otherwise walk or cycle).

Steve Munro is probably not the only person who is confused about the Metrolinx BCAs. It is hard to know what to make of them when Metrolinx itself ignores their findings and proceeds with projects that show a very poor ratio of costs to benefits.

It is not enough simply to commission a BCA; if Metrolinx intends to deliver on its promises of investing "where it matters most," prudent financial management, accountability, and risk management, it should actually heed the BCAs if they show that a project makes no financial sense.

Michael Schabas