Determination No. R-2020-194

SUMMARY

[1] This is the Agency’s determination of the 2021 regulated interswitching rates pursuant to Part III, Division IV of the CTA. It builds on the methodology set out in Determination No. R-2018-254 (2019 interswitching rates Determination) and Determination No. R-2019-230 (2020 interswitching rates Determination), and it contains determinations on two issues that were identified in and were the subject of a comprehensive consultation, which was launched on June 20, 2019: Item 5 – Contribution to fixed costs and Item 6 – Productivity factors.

[2] The Agency will address the following issues:

1. Should there be any changes to the methodology used to calculate the Canadian National Railway Company’s (CN) and the Canadian Pacific Railway Company’s (CP) contributions to fixed costs in the development of the regulated interswitching rates?
2. Should there be any changes to the methodology used to calculate CN’s and CP’s productivity growth factor in the development of the regulated interswitching rates?
3. What are the 2021 regulated interswitching rates?

[3] With respect to the calculation of CN’s and CP’s contributions to fixed costs and their productivity growth factor in the development of the regulated interswitching rates, the Agency reaffirms the methodology set out in the 2019 interswitching rates Determination for the reasons set out below.

[4] The methodology used by the Agency in the determination of the 2021 interswitching rates is presented in Appendix A.

[5] The Agency determines the regulated interswitching rates for 2021 under subsection 127.1(1) of the CTA as follows:

Schedule of regulated interswitching rates for 2021

Item	Column I – Interswitching distance zone	Column II – Rate per car for interswitching traffic to or from a siding (single car)	Column III – Rate per car for interswitching a car block (60 cars or more)
1	Zone 1	$290	$80
2	Zone 2	$405	$125
3	Zone 3	$310	$70
4	Zone 4A	$260	$100
4	Zone 4B	$260 + $8.50 per additional km	$100 + $1.05 per additional km

BACKGROUND

[6] Regulated in Canada since 1904, interswitching is part of the competitive access provisions that give some shippers access to the services of railway companies that do not directly serve their facilities or sidings. The interswitching provisions require a railway company that does provide such direct service to transfer cars with a shipper’s traffic at an interchange to a different railway company with which the shipper has made transportation arrangements. The transportation to the interchange must be done at a prescribed rate. The Agency is responsible for calculating and publishing that rate.

[7] Under the CTA, the Agency must have regard to certain considerations in setting the rate, including the following:

• Section 112 requires the rate to be commercially fair and reasonable to all parties.
• Paragraph 127.1(2)(a) requires the Agency to take into consideration any reduction in costs that, in the Agency’s opinion, results from moving a greater number of cars or from transferring several cars at the same time.
• Subsection 127.1(3) requires the Agency to consider the average variable costs of all movements of traffic that are subject to the rate, and that the rate shall not be less than the variable costs of moving the traffic, as determined by the Agency.

[8] The Transportation Modernization Act, SC 2018, c 10 (Transportation Modernization Act), which received royal assent on May 23, 2018, made several amendments to the regulated interswitching provisions in the CTA. Paragraph 127.1(2)(b) of the amended CTA now requires the Agency to also take into consideration any long-term investment needed in the railways. The following amendments were also made:

• Subsection 127.1(1) requires the Agency to determine the interswitching rate no later than December 1 of every year.
• Subsection 127.1(4) requires the Agency to publish the method that it followed for determining the rate.
• Section 128.1 requires the railway companies to provide to the Agency the information or documents that the Agency considers necessary to exercise its powers or perform its duties or functions under section 127.1.

[9] Prior to these amendments, updates to interswitching rates were done by regulation, which resulted in a relatively significant lag time between updates. The shift to annual updates ensures that the rates are up to date and fully compensatory.

[10] The Agency launched a comprehensive consultation on June 20, 2019, asking stakeholders to comment on a broad range of methodological questions and factors with potential impacts on the interswitching rate methodology and on the Railway Interswitching Regulations, SOR/88-41 (Interswitching Regulations). The deadline for written submissions on this consultation was August 21, 2019.

[11] In the discussion paper for the 2019 consultation (discussion paper), the stakeholders were asked to comment on a total of 18 questions related to the following 9 items:

• Item 1: Federally regulated short-line railway companies;
• Item 2: Regional and commodity-specific regulated interswitching rates;
• Item 3: Interswitching zones up to 30 km;
• Item 4: Long-term investment needs of the railway companies (cost of capital methodology);
• Item 5: Contribution to fixed costs;
• Item 6: Productivity factors;
• Item 7: Volume discount rate categories;
• Item 8: Collecting interswitching service units; and
• Item 9: Transparency of the regulated interswitching rates and methodology.

[12] The stakeholders were also offered the opportunity to raise any other issues that they would like to discuss.

[13] In the 2020 interswitching rates Determination and in Determination No. R-2019-229, the Agency made determinations on these items, with the exception of Items 5 and 6.

[14] On August 1, 2019, the deadline for written submissions with regard to Items 5 and 6 was extended until January 25, 2020.

[15] Having now reviewed Items 5 and 6 in this Determination, the Agency has completed the full review of its interswitching rate methodology that it began in 2019 following the coming into force of the Transportation Modernization Act.

ISSUE 1: SHOULD THERE BE ANY CHANGES TO THE METHODOLOGY USED TO CALCULATE CN’S AND CP’S CONTRIBUTIONS TO FIXED COSTS IN THE DEVELOPMENT OF THE REGULATED INTERSWITCHING RATES?

Description

[16] In the discussion paper, the Agency sought input on whether it should continue to use its current methodology or an alternative methodology, such as the full Ramsey pricing model (Ramsey pricing), to calculate the contribution to fixed costs.

[17] The Agency currently applies a contribution to fixed costs on top of the system average variable costs to calculate the total economic costs of a railway company at the system-wide level. This contribution is calculated separately for each railway company, where their total system cost (which is obtained from their annual report) is divided by the system variable cost that is calculated using the Agency’s costing model.

[18] As shown in Appendix B, the interswitching rates calculation includes a fixed costs factor that accounts for the costs that a railway company incurs during its operations, independent of the freight movement. The fixed costs of a railway company represent expenses that are fully independent of traffic size, such as the maintenance of bridges and snow removal, as well as the non-variable portion of expenses in accounts that do not vary fully with traffic. In the majority of cases, they are costs incurred by the railway companies that are not linked to any specific movement, but serve to benefit all movements across their system.

[19] As the Agency explained in the 2019 interswitching rates Determination, the total economic cost provided in the interswitching rates ensures that the railway companies are financially viable in the long-term by addressing the requirement that the rates are commercially fair and reasonable (section 112 of the CTA), and also addresses their long-term investment needs as required by paragraph 127.1(2)(b) of the CTA.

[20] In economic literature ¹, Ramsey pricing is known as the second-best solution concerning what prices a monopoly should charge for the products that it sells in order to maximize social welfare while earning enough revenue to cover its fixed costs.

[21] In the case of a railway company, the first-best solution where it sets the price of each product to be equal to its marginal cost would result in the railway company going out of business due to the considerable fixed costs as is typical with the railway industry. For a railway company with large fixed costs, setting the price equal to the marginal cost would result in a price less than average cost, and the railway company would not survive without subsidies.

[22] Ramsey pricing aims to maximize social welfare by optimizing the resources of the economy such that these resources cannot be reallocated to benefit someone without making someone else worse off. Ramsey pricing proposes to charge a seller’s differential prices for the same product based on their demand elasticity. Ramsey pricing usually has a limit set to the maximum revenue that the seller is allowed to earn.

Consultation feedback

[23] There were no participants who supported the use of Ramsey pricing to calculate the contribution to fixed costs. However, certain participants shared some additional opinions on the Agency’s current approach to the contribution to fixed costs.

[24] CN is of the opinion that the contribution to fixed costs should be at least equal to, and likely higher than, the system average contribution. CN argues that interswitching activities are of a shorter distance than the average line haul movements, and because fixed costs are independent of the distance, this suggests that the contribution to fixed costs for interswitching should be set above the system average contribution to fixed costs.

[25] CP argues that the Agency’s regulatory costing model is not compensatory due to data issues in determining a railway company’s variable costs. Due to these data issues, CP claims that the Agency’s costing model will include errors and have implications on CP’s recovery of interswitching costs:

Accurate costing requires that costs are assigned, to the best ability of the [Agency], where they accrue. This outcome is predicated on the robustness of the statistical methods and data used to derive the variabilities. However, there are data issues that are a cause for concern.

CP highlights that it is modernizing its data systems and leveraging new technologies to better allocate cost and operating records. These changes have produced gradual changes in certain accounts. The concern is that long term regressions based on data dating back to 1992 will include errors as a result of the variation in accounts caused directly by data modernization.

[26] Teck Resources Limited, the Western Grain Elevator Association, the Canadian Canola Growers Association, the Mining Association of Canada, and the Western Canadian Shippers Coalition, all represented by McMillan LLP (McMillan), believe that the current markup of a minimum, or lower than the system average, contribution to fixed costs is a reasonable and reliable approach. McMillan suggests that the Agency should continue using the present methodology as it seems to adequately compensate the railway companies.

[27] The Forest Products Association of Canada (FPAC) believes that the current methodology that is used by the Agency is appropriate and that Ramsey pricing should not be used; however, it did not provide a rationale for its response.

Analysis and determinations

[28] No participant was in favour of Ramsey pricing.

[29] As stated above, Ramsey pricing is recognized in economic literature as the second-best solution in a monopolistic industry to maximize social welfare, that is, in this context, the welfare of both the railway companies and the shippers. The first-best solution is not viable for an industry with large fixed costs as businesses would go out of business. However, a primary concern or challenge with Ramsey pricing, as submitted by McMillan, is the practical estimation of demand elasticity. It is difficult to obtain data on the demand behaviour of consumers when prices change.

[30] Ramsey pricing in the context of railway transportation is further complicated by the fact that the demand for railway services is a derived demand. Using the example of coal shippers from McMillan’s submission, a coal shipper’s demand for railway service is dependent on the demand for coal in the economy at any given time and not the price of moving coal through rail.

[31] The Agency’s current methodology to calculate the contribution to fixed costs is an approximation of Ramsey pricing where the elasticity of all shippers is assumed to be the same, and they are therefore charged the same price per zone. The contribution to fixed costs is set at a level where, if all shippers were charged the same contribution to fixed costs, then the railway company would earn enough revenue to cover its fixed costs.

[32] This approach is easy to understand and uses reliable data to calculate the contribution to fixed costs. Ramsey pricing would require data on how each shipper’s demand reacts to fluctuations in the freight railway service prices. This data would then be used to develop shippers’ demand elasticities. However, data of shippers’ demand sensitivity to freight service prices is not available from any reliable source.

[33] Given the difficulties with and the lack of support from participants for Ramsey pricing, the Agency will continue to calculate the contribution to fixed costs using the system average total costs to variable costs ratio based on the railway companies’ submitted data to ensure that the interswitching rates determined by the Agency are fair and reasonable to all parties.

[34] With respect to CN’s proposal, if CN is suggesting that the contribution to fixed costs of its non-regulated movements is less than the contribution to fixed cost calculated by the Agency, then it should have provided this evidence to the Agency.

[35] CN’s submission dated August 21, 2019, provides some insight that its non-regulated movements do not have a lower contribution to fixed costs than that calculated by the Agency as CN submits that it has earned approximately 16 percent on invested capital annually over the last 3 years, versus the Agency’s allowed cost of capital of 5 percent.

[36] As shown in Appendix B, the calculation for the contribution to fixed costs is based on a mathematical calculation such that, so long as every shipper is charged the same percentage markup, the railway companies will recover their full economic costs.

[37] However, based on CN’s submission dated August 21, 2019, the contribution to fixed costs arising from its non-regulated movements is more than what is calculated by the Agency, resulting in the generation of more funds than are necessary to cover its full economic costs. The suggestion by CN that interswitching movements should form a larger contribution of fixed costs would lead to an even larger contribution above total economic costs.

[38] Generating more funds than are necessary to cover the railway company’s full economic costs will result in shareholders receiving a return on investment that is greater than the risks assumed for the period under consideration. It is the Agency’s opinion that for the interswitching rates to be fair and reasonable to all parties as required by section 112 of the CTA, the interswitching rates should cover the full economic costs of the railway companies and provide a rate of return to shareholders that is consistent with the risks assumed for the period under consideration. The Agency conducted a full review of its cost of equity model last year and determined a methodology in Determination No. R-2019-229 to ensure that it is providing the return that shareholders require. Coupled with the fact that the Agency is providing full economic costs based on the railway companies’ submitted data, the Agency has satisfied the requirements of section 112 of the CTA and, therefore, rejects CN’s proposal.

[39] With respect to CP’s submission, the Agency’s regulatory costing model relies on data that is submitted by the railway companies. As a result, if any data issues exist, they would stem from the submissions of the railway company and are not due to the Agency’s regulatory costing model.

[40] If CP is aware that its data recording processes in the past are producing inaccurate results and it is currently working on improving those processes, it could provide annual cross-sectional data to address its concerns.

[41] In addition, as stated above, the Agency provides a contribution to fixed costs on top of the variable costs in the calculation of interswitching rates, which results in the railway companies recovering their total economic costs of providing interswitching service. It is unclear to the Agency why CP believes that there are implications on CP’s recovery of interswitching costs due to its potential data issues when the Agency provides total economic costs in the regulated interswitching rates.

[42] The Agency is using the most up-to-date information available to it, and welcomes CP submitting new information for next year’s rate assessment.

[43] Given that CP has provided no evidence to support its assertion that the Agency’s rates are non-compensatory due to errors in calculating the variable costs associated with regulated interswitching nor provided any alternative data to be considered, the Agency will maintain its current methodology for determining the railway companies’ variable and fixed costs.

ISSUE 2: SHOULD THERE BE ANY CHANGES TO THE METHODOLOGY USED TO CALCULATE CN’S AND CP’S PRODUCTIVITY GROWTH FACTOR IN THE DEVELOPMENT OF THE REGULATED INTERSWITCHING RATES?

Description

[44] In the discussion paper, the participants were asked if there is an alternative way to calculate the railway companies’ productivity growth factor (productivity) that does not assume perfect competition or constant returns to scale, such as the Malmquist Productivity Index.

[45] The interswitching rates developed by the Agency are forward-looking rates based on historical railway company unit costs. In order to arrive at estimated costs for the upcoming year, the Agency must apply inflation and productivity to historical unit costs.

[46] Productivity is a measure of how efficiently a railway company uses its inputs (labour, capital, material, etc.) to produce outputs (various railway services) when compared between two time periods. If a railway company uses fewer resources for the same volume of output, or if it produces a higher quantity of outputs using the same level of inputs, then it implies that this railway company has gone through a productivity growth phase. A railway company could also experience a decline in its productivity growth if more resources are used to produce the same volume of output, or if less output is produced using the same level of inputs compared to a different year.

[47] The productivity of a railway company can be estimated either as a single factor productivity growth measurement, which examines total outputs with a single input, or a total factor productivity (TFP) growth measurement, which compares total outputs with total inputs.

[48] TFP is one of the primary ways that economists estimate productivity because of its broader scope. The current methodology used by the Agency to estimate TFP is the Ideal Fisher methodology, where the aggregate growth rate of outputs index is compared to the aggregate growth rate of inputs index. Productivity is calculated using the data on quantities and prices submitted by CN and CP annually to the Minister of Transport in various schedules. The prices for inputs such as labour, material, equipment and land come partly from the price indices for each railway company that are developed internally by the Agency. Information on the various schedules and data used is presented in more detail in Appendix A.

[49] The Ideal Fisher methodology is based on an academic paper written by W.E. Diewert in 1992. The underlying assumptions of the Ideal Fisher methodology are that there is perfect competition among the railway companies and there are constant returns to scale for those railway companies, that is, a change in output is proportional to the change in inputs.

[50] Perfect competition assumes that a railway company has no influence on the price that it charges to shippers for the services that they offer (that is, the railway company has no market power). The second assumption of constant returns to scale implies that an increase in the size of the railway company does not affect its efficiency gains.

[51] Notwithstanding these assumptions, the Ideal Fisher methodology is a widely used methodology for assessing productivity due to its simplicity and ease of calculation, including by the Surface Transportation Board to calculate the productivity of railway companies in the United States of America.

Consultation feedback

[52] CN expressed concerns about the negative impact that a productivity adjustment has on interswitching rates. CN submits that an investment by a railway company to increase its efficiency, which results in a decrease in interswitching rates, could act to discourage productivity growth. CN further submits that increases in productivity should result in an increase on the return on investment.

[53] CN indicates that it does not have a preference with respect to either productivity model; however, it submits that any model used should have an adjustment to account for any under- or over-calculation of the interswitching rates as a result of the forecasting errors stemming from the Agency’s methodology. It is CN’s opinion that there are forecasting errors in the Agency’s interswitching rates calculation, which are underestimating costs, such that the Agency is failing to meet the requirements of subsection 127.1(3) of the CTA.

[54] CP submits that the Agency should no longer apply a productivity adjustment in its interswitching rates calculation, as a productivity adjustment removes the incentive to innovate. CP also submits that should productivity continue to be applied, then its scope should be narrowed to only apply to interswitching operations and not system-wide productivity. CP believes that a railway company’s productivity has been found in railway operations apart from interswitching activities, such as running longer trains and improving fuel efficiency. CP submits that interswitching has likely contributed very little to the increase in the railway companies’ productivity.

[55] FPAC believes that the Ideal Fisher methodology that the Agency currently uses is the best method to calculate a railway company’s productivity; however, it did not provide a rationale for its response.

Analysis and determinations

[56] With respect to the comments made about the disincentive to invest, as noted in the 2020 interswitching rates Determination, the railway companies have a service-level obligation to carry all traffic on offer and must furnish a level of service for interswitching that is equal to the level of service for line haul traffic. Railway companies are expected to make the necessary investments in infrastructure supporting regulated interswitching to ensure that they can meet these legal requirements.

[57] The assumptions of the Ideal Fisher methodology do not entirely hold true in the context of a Canadian Class I railway company. First, perfect competition may not be a realistic assumption for a Canadian railway company given that portions of the market are only serviced by one railway company, and for some shippers and the commodities that they ship, other transportation modes are not a competitive option to railway transportation.

[58] Secondly, the assumption of constant returns to scale does not hold true in general for the railway industry given that there are considerable fixed costs involved. These fixed costs must be put into place even if only one unit of output is produced, and must be paid even with zero output. In this case, the larger the output, the more these fixed costs can be spread out, resulting in economics of scale, and not constant returns to scale.

[59] The Malmquist Productivity Index does not assume perfect competition or constant returns to scale. However, this calculation is more complex and may, therefore, not be as transparent as the Ideal Fisher methodology.

[60] In an Agency research paper provided to participants, the Agency assessed the 2011 and 2012 interswitching rates based on applying the Ideal Fisher methodology and the Malmquist Productivity Index. The Agency has updated the research paper previously provided to participants and found that when it applied inflation and productivity to a hypothetical rate of $300 in 2010, the interswitching rates produced in 2011 would be $325 according to the Ideal Fisher methodology and $330 according to the Malmquist Productivity Index, while the rates produced in 2012 according to both methodologies would be the same, $295.

[61] The above analysis shows that the two models do not produce significantly different results, yet the Malmquist Productivity Index is much more difficult to apply.

[62] According to the Organisation for Economic Co-operation and Development, the Ideal Fisher methodology is a superlative index, which it defines as:

Superlative indices are price or quantity indices that are ‘exact’ for a flexible aggregator. A flexible aggregator is a second-order approximation to an arbitrary production, cost, utility or distance function. Exactness implies that a particular index number can be directly derived from a specific flexible aggregator.

[63] The measurement of productivity for a railway company is complicated because the output is a derived demand. In addition, measuring the productivity for only the interswitching portion of the railway services further complicates the process. Given that the interswitching rates are now calculated annually, it is essential to keep the productivity growth estimation process as simple and transparent as possible.

[64] The Agency finds that, given the similar results that the Ideal Fisher methodology and the Malmquist Productivity Index produced above, the assumptions of perfect competition and constant returns to scale are a flexible aggregator and can be estimated using the Ideal Fisher methodology. No participants in the consultation expressed any concern with the Ideal Fisher methodology nor expressed any support for the Malmquist Productivity Index. In the circumstances, the Agency finds that it is appropriate to continue applying the Ideal Fisher methodology.

[65] With respect to CN’s request for a forecasting error adjustment, a forecasting error is only a concern if it is consistently one-sided and demonstrable. As the CTA was amended to direct the Agency to calculate interswitching rates on an annual basis, sufficient evidence is not available to assess the magnitude of the annual forecasting error, or if this error eventually balances out to zero over a given time period.

[66] In addition, the Agency is concerned with the application of such an adjustment (if an adjustment is determined to be required). The CTA permits the Agency to determine the rate per car to be charged for the following calendar year as provided for in subsection 127.1(1). If a correction mechanism was put in place to adjust for forecasting errors, this correction mechanism may also lead to additional errors as the Agency can only include the correction mechanism in the rate per car. For example, if it was determined that the 2019 regulated interswitching rates had a forecasting error of $100, and in 2019, the number of cars moved under regulated interswitching was 100, then the adjustment would be $1 per car for the following year’s interswitching rates determination. However, this adjustment would be based on 2019 carloads. If the following year (2020 in the example)—when this adjustment is applied—evidences usage of regulated interswitching of more than 100 cars, this would result in an overpayment of the error adjustment and the need for another adjustment to be made in the subsequent year’s interswitching rates determination. This is further complicated by the fluctuating share of traffic, from year to year, that each railway company moves.

[67] As there is no evidence to suggest the existence of forecasting errors, the Agency will not implement any correction mechanism at this time, but will assess whether it is required as evidence accumulates over time.

ISSUE 3: WHAT ARE THE 2021 REGULATED INTERSWITCHING RATES?

[68] The calculation of the 2021 regulated interswitching rates relies on available data, and uses well-established costing methodologies, some elements of which are used in other Agency determinations. It also reflects relevant methodological determinations, including:

• Determination No. R-2019-229 – Review of the methodology used by the Agency to determine the cost rate of common equity for federally-regulated railway companies, dated November 29, 2019;
• Determination No. R-2019-230 – Determination by the Agency of the 2020 regulated interswitching rates pursuant to Part III, Division IV of the CTA, dated November 29, 2019;
• Determination No. R-2017-198 – Determination by the Agency of the methodology to be used by federally-regulated railway companies to determine the working capital amounts and capital structure for regulatory purposes, dated December 5, 2017;
• Order No. 2015-R-91 – Determination by the Agency of the variable portions of railway company cost accounts for CN and CP, dated June 8, 2015; and
• Decision No. 425-R-2011 – Review of the methodology used by the Agency to determine the cost of capital for federally-regulated railway companies, dated December 9, 2011.

[69] The data used in the development of the interswitching rates are as follows (Appendix A describes these components in greater detail):

1. Interswitching service units (obtained through conference calls with complementary electronic data from CN and CP);
2. 2017 unit cost for CN and 2018 unit costs for CP for each service unit, including overheads (approved by the Agency on November 29, 2019 and November 30, 2020 respectively);
3. Contribution to fixed costs (the data required for this calculation is from CN’s 2017 and CP’s 2018 annual reports to the Minister of Transport and was obtained by the Agency through Transport Canada on October 24, 2019);
4. 2021 forecasted component costs for CN and CP (obtained through the Agency’s calculation of the 2020-2021 Volume-Related Composite Price Indices in Determination No. R-2020-81);
5. 2021 cost of capital rate for regulated interswitching (data obtained pursuant to Determination No. R-2019-229, the Determination of the methodology to determine the working capital amounts and capital structure for regulatory purposes, and the 2011 cost of capital Decision);
6. 2021 productivity rates for CN and CP (data from CN’s and CP’s annual reports to the Minister of Transport from 2017-2018, and various tables from Statistics Canada); and
7. 2019 volumes of interswitched cars (submitted by CN on April 3, 2020, and by CP on April 9, 2020).

[70] The current approach to determining interswitching rates is based on actual service units within each zone. Costs are affected by a range of factors that can include train length, customer siding characteristics, and train yard activities, any of which can vary considerably from one situation to another.

2021 REGULATED INTERSWITCHING RATES

[71] Based on the determinations made after the consultation process and on the application of the methodology outlined in Appendix A on the data, the Agency determines the regulated interswitching rates for 2021 in the following schedule, according to the interswitching distance zones and car block as defined in the Interswitching Regulations:

Item	Column I – Interswitching distance zone	Column II – Rate per car for interswitching traffic to or from a siding (single car)	Column III – Rate per car for interswitching a car block (60 cars or more)
1	Zone 1	$290	$80
2	Zone 2	$405	$125
3	Zone 3	$310	$70
4	Zone 4A	$260	$100
4	Zone 4B	$260 + $8.50 per additional km	$100 + $1.05 per additional km

[72] Where a siding is located in Zone 4B, the interswitching rate for each car is increased from Zone 4A for each kilometre over 40 km by $8.50 per car for single-car movements or by $1.05 per car for car-block movements.

[73] Any required additional kilometres are calculated by identifying the shortest distance, along the line of track of a terminal carrier, from an interchange to the point of connection with the siding.

[74] For all other zones, the interswitching rate charged by a terminal carrier for traffic originating in, or destined to, an interswitching distance zone set out in Column I of the schedule is the interswitching rate set out in Column II or III, as the case may be.

[75] For the movement of intermodal containers, the rate per car is based on the number of platforms, which is the most comparable traffic unit for localized intermodal rate determination purposes.

---

APPENDIX A

The 2021 interswitching rates calculated by the Agency are based on a methodology that captures the economic costs of providing interswitching services. These economic costs include explicit costs such as operating costs, including the depreciation of assets, as well as the implicit costs associated with the returns on investment in those assets. The returns on investment are a weighted average of the returns on debt and the returns on equity, and are determined by the Agency according to its cost of capital methodology based on Decision No. 425-R-2011 (2011 Decision), Determination No. R-2017-198 and Determination No. R-2019-229 (2019 Determination).

For explanatory purposes, the Agency has calculated interswitching rates based on the following simplified formula:

Interswitching rates$(A\times B)\times \genfrac{}{}{0.1ex}{}{C}{D}$

Where:

• A is interswitching variable costs
• B is contribution to fixed costs
• C is a factor to account for price inflation, and
• D is a productivity adjustment factor

Interswitching variable costs$\left(A\right)=(\genfrac{}{}{0.1ex}{}{E}{F}\times G)\times H$

• E is system costs
• F is system service units
• G is variability of costs, and
• H is interswitching service units

The expression $(\genfrac{}{}{0.1ex}{}{E}{F}\times G)$ is referred to below as the unit cost for each service unit, including overhead.

A more detailed explanation can be found in Appendix B. In the following sections each of these variables is described in further detail.

1.0 Interswitching service units

Every year, Agency staff visits interchange locations across Canada to meet with CN and CP yard supervisors to review interswitching operations at each location. For each interchange location, all of the steps required to provide interswitching services for the major interswitching shippers in each zone and to estimate the service units involved in each step are verified. Agency staff visits interchanges of different sizes, volumes and characteristics to capture the unique operations of interchanges across Canada. Over a two-year period, Agency staff will update service units from all interchanges that are providing regulated interswitching service.

The Agency typically determines interswitching service units through conducting a combination of annual staff site visits and conference calls with complementary electronic data from CN and CP.

For 2020, due to the COVID-19 pandemic, Agency staff collected all data regarding interswitching service units via conference calls to protect the health of its staff, staff from CN and CP, and any Canadians who would be in contact with Agency or railway company staff during travel associated with in-person interswitching site visits. The Agency expects to return to a combination of targeted in-person interswitching site visits and conference calls as soon as it is safe to do so.

The service units determined for single-car rates and block-train rates are described in further detail in sections 1.1 and 1.2 respectively.

1.1 Single-car service units

There are two different types of interswitching operations for single-car movements (interswitching 59 cars or less):

• Yard switching; and
• Road switching.

Under yard switching, a yard crew will pick up the interchange cars at the interchange and will bring them back to the yard for classification (sorting) and marshalling (placing cars in order for delivery). Cars are then delivered to the customer. On the return trip, the cars are returned to the yard where they are classified and marshalled again before returning to the interchange.

Road switching occurs in locations where switching in a yard is not possible, or in situations where only minimal classification or marshalling is required. Road switching involves either a line-haul train or a road crew picking up cars at the interchange. The cars may or may not be classified or marshalled at the interchange before being delivered to the customer. On the return trip, the cars are brought back to the interchange with little or no classification or marshalling.

Service units determined for road switching include:

• Gross ton-miles – which drive costs such as track maintenance;
• Car-miles – which drive costs such as car inspection;
• Train-miles – which drive costs such as signals maintenance;
• Carloads – which drive costs such as marketing and sales;
• Fuel consumed;
• Crew wages; and
• Diesel unit miles – which drive costs such as locomotive maintenance and investment.

Yard switching is more complex in terms of classification and marshalling. In most major yards, there would be a dedicated yard assignment with crews classifying or marshalling hundreds of cars. As tracking specific cars and mileage at the yard is not possible in all circumstances, mileage at the yard is simplified as yard-switching minutes.

Yard-switching minutes capture the amount of time that it takes to service a customer, including the process of classification and marshalling. The associated unit cost for this service unit captures all of the expenses incurred for yard switching, including crew wages, locomotive fuel expenses, locomotive maintenance expenses, and track and roadway maintenance.

Service units for single-car movements increased on average by 5.22 percent compared to the service units that were used in Determination No. R-2019-230, translating to an average increase of $14.54 per car in the interswitching rates.

This increase was driven by new safety protocols in CN’s Thornton yard resulting in a decrease in the number of cars that it can process per day. In addition, railway company staff provided more detailed information on the unproductive time that their operations crew encounters during the course of their shift, including delays at bridges, safety briefings with shippers with dangerous goods, the beginning of shift preparation times and the end of shift reporting procedures.

1.2 Block-train service units

Service units determined for block trains include:

• Gross ton-miles;
• Car-miles;
• Train-miles;
• Carloads;
• Fuel consumed;
• Crew wages; and
• Diesel unit miles.

Block movements involve a “hook and haul” operation where blocks of cars are hooked on at the interchange and delivered directly to the customer. On the return trip, cars are hooked and delivered directly to the customer. However, additional handling, either at the interchange or at the shipper siding, may be required. If, for example, the siding or the interchange is not long enough to handle the block, the railway company must perform one or multiple cuts to the block in order to complete the movement. Where additional handling is identified during site visits, the costs are reflected in the final interswitching rate.

For block-car operations, service units increased by 33.14 percent compared to the service units that were used in Determination No. R-2019-230, translating to an average increase of $14.30 per car in the interswitching rates.

This increase was driven by three new grain shippers in the Vancouver area that are now utilizing regulated interswitching. These shippers’ average train length is shorter than that of existing shippers, resulting in larger service units in the calculation of the weighted average rates.

2.0 Unit cost for each service unit, including overheads

Derived service units are multiplied by their corresponding unit cost to obtain a cost per car for each shipper in each zone. CN and CP submit their detailed financial and operating data to the Agency each year based on the Agency’s Uniform Classification of Accounts And Related Railway Records (2014) [UCA]. The UCA defines the method of accounting for railway companies subject to regulation by the Agency. It provides accounting instructions and the framework of accounts for the rail operations of such railway companies. It also provides instructions for recording operating statistics and defines the categories for these data.

The Agency approves each railway company’s cost to produce a unit of defined railway activities such as track and roadway maintenance, signals investment, and the like, based on system expenses for each activity and the observed system service units.

The costing model developed by the Agency then determines the total variable cost, including direct activities as well as indirect supervisory, management and administration activities, to produce a unit cost for each service unit. These indirect costs are referred to as overhead, as they do not relate to service units directly, but instead relate to the direct costs of those service units (for example, when a train moves one gross ton mile, it will incur track maintenance labour costs directly, as well as indirect costs or overhead for the management of, and the equipment used by, track maintenance labourers).

For the 2021 interswitching rates, the Agency has used the 2017 unit costs for CN and the 2018 unit costs for CP. An index factor (using indices from the volume-related composite price index) and a productivity factor are applied (based on the Agency’s current productivity model ²) in order to estimate costs in 2021. The equation for the Agency’s productivity model is presented in Appendix B.

The data used in the productivity model are the following:

Output price and quantity data

CP’s output data

Price (index 2017 = 1)	Price = freight revenue / revenue ton miles
Output quantity	Revenue ton miles CP output data is categorized into 7 commodities: Grain and grain products Mine products Agriculture products Animal products Forest products Intermodal Manufactured and misc. products

Source : Output quantity and price from F47 submitted annually by CP to the Agency

CN’s output data

Price (index 2017 = 1)	Price = freight revenue / revenue ton miles
Output quantity	Revenue ton miles CN output data is categorized into 13 commodities: Agricultural products Grain Coal Forest products Paper and pulp Fertilizers Mineral ores Metals Automobiles and parts Fuel and chemicals Mine products Manufactured products Intermodal

Source : Output quantity and price from S40 submitted annually by CN to the Agency

Input price and quantity data

Labour

Price (index 2017 = 1)	Labour price (labour price = annual labour price per hours / annual labour price per hour of base year)
Quantity	Hours worked (hours worked = labour expense / labour price)

Source : Schedule 12 submitted by CN and CP

Fuel

Price (index 2017 = 1)	Fuel price (fuel price = annual fuel price per litre / annual fuel price per litre of base year)
Quantity	Litres consumed (litres consumed = fuel expense / fuel price)

Source : Schedule 13 submitted by CN and CP

Material

Price (index 2017 = 1)	Material price index (MPI) (MPI = Agency calculated MPI for a year / Agency calculated MPI of base year)
Quantity	Material quantity (material quantity = material expense / MPI)

Source : F-46, Schedule 12 and Schedule 13 submitted by CN and CP

Land

Price (index 2017 = 1)	Ptland (please refer to Note 1)
Quantity	Material quantity (material quantity = material expense / MPI)

Source : F-46, Schedule 12 and Schedule 13 submitted by CN and CP

Way and structure capital

Price (index 2017 = 1)	Ptw&s (please refer to Note 1)
Quantity	Annual net book value of way and structure

Source : Way and structure quantity from F-49

Owned equipment

Price (index 2017 = 1)	Ptowned_eqp(please refer to Note 1)
Quantity	Annual net book value of owned equipment

Source : Owned equipment quantity from F-49

Leased equipment

Price (index 2017 = 1)	Ptleased_eqp (please refer to Note 1)
Quantity	Leased equipment expense

Source : Leased equipment quantity from F-46 (551-566)

Note 1

Service price for land (Ptland)=11-ut1+coctAt-1-At+StAt

Service price for way and structure (Ptw&s)=1-utdt1-ut1+coctAt-1-1-δtAt+StAt

Service price for owned equipment (Ptowned_eqp)=1-kt 1-utdt1-ut1+coctAt-1-1-δtAt+StAt

Service price for leased equipment (Ptleased_eqp)=1-kt 1-ut1+coctAt-1-At+StAt

Where:

• kt is the investment tax credit rate;
• ut is the marginal corporate income tax rate;
• At is the asset price;
• dt is the present value of all future depreciable deductions for tax purposes;
• coct is the annual cost of capital rate as determined by the Agency;
• δt is the annual replacement rate; and
• St is the property tax rate.

Assumption:

• The investment tax credit rate (kt) is set to zero.

Data sources:

• The marginal corporate income tax rate ut is submitted annually to the Agency by CN and CP.
• The asset price At is from Statistics Canada: land (Table 18-10-0205-01); way and structure, owned equipment, leased equipment (Table 36-10-0097-01).
• (dt)is from schedule F-49 that is submitted annually by CN and CP to Transport Canada.
• (St)is from schedules F-46 and F-49 that are submitted annually by CN and CP to Transport Canada.

The Agency applied an average productivity value of 86.27% as the estimated productivity gain for 2021. This is based on the Agency’s calculation of the average total factor productivity growth of each railway company from 2017 to 2018.

The productivity value increased compared to the value that was used in Determination No. R-2019-230, which is a result of the Agency deciding, in this Determination, to resume the calculation of CN’s and CP’s productivity value on an annual basis. In previous determinations, the Agency was using a 10-year average of productivity values from when it last calculated the productivity.

The table in Appendix C lists all of the variable cost accounts (as defined by the UCA) that factor into the 2021 interswitching rates.

3.0 Cost of capital

The 2011 Decision sets out that with the exception of the risk-free rates of return, all the elements that are necessary for the calculation of the cost of capital rate for the purposes of establishing regulated interswitching rates will be those determined annually in the cost of capital rate for the transportation of western grain.

With respect to the appropriate risk-free rates, the 2019 Determination sets out that for the cost of capital rate for the purposes of establishing regulated interswitching rates:

• the Canadian risk-free rate will be the average yield on Government of Canada 5-10 year marketable bonds for the month of September of the year prior to the interswitching year, as published by the Bank of Canada; and
• the U.S. risk-free rates will be the average yields on each of 5-year and 10-year U.S. Treasury bonds for the month of September of the year prior to the interswitching year, as published by the U.S. Federal Reserve.

Based on this, the Canadian risk-free rate of 0.48 percent, the U.S. 5-year risk-free rate of 0.27 percent, and the U.S. 10-year risk-free rate of 0.68 percent were used for the calculation of the 2021 interswitching rates.

The resulting cost of capital rate used in the development of the 2021 interswitching rates is 4.64 percent for CN and 4.60 percent for CP.

The September 2019 risk-free rates used in last year’s interswitching cost of capital rates were 1.37 percent for Canada, 1.57 percent for the U.S. 5-year risk-free rate and 1.70 percent for the U.S. 10-year risk-free rate.

The significant decline in the risk-free rates was due to the decline of the global economy caused by the COVID-19 pandemic and global trade wars.

4.0 Volumes of interswitched cars

The volumes of interswitched cars are required to calculate a weighted system average of costs starting at each interchange, then for each zone, and finally for CN and CP, to come up with the aggregated weighted system average interswitching costs. The hypothetical example below illustrates the weighting that is applied:

• For each interchange, the costs per car for each shipper within a zone are weighted by the carloads interswitched to produce an average cost per interchange.

Table 1: Calculating costs for Vancouver zone 1 interswitching for railway company ABC

Vancouver zone 1 for railway company ABC	2017 carloads	% weight (share of traffic)	Variable cost per car	Weighted zone 1 cost
Shipper A	800	80%	$100	$80
Shipper B	200	20%	$80	$16.00
Vancouver zone 1 cost per car for railway company ABC				$96

• For each zone, the average costs for each interchange are then weighted by the traffic interswitched to produce an average cost per car for each zone. For example, the result from Table 1 is found in the first row below.

Table 2: Calculating zone 1 interswitching costs for railway company ABC

Zone 1 for railway company ABC	2017 carloads	% weight (share of traffic)	Variable cost per car	Weighted zone 1 cost
Vancouver	1,000	62.50%	$96	$60
Toronto	600	37.50%	$150	$56.25
Zone 1 cost per car for railway company ABC				$116.25

• The costs for each railway company in each zone are then averaged, based on the interswitching traffic of each railway company in that particular zone, to generate a system average variable cost measure per car for each of the four distance zones. For example, the result from Table 2 is found in the first row of Table 3.

Table 3: Calculating zone 1 interswitching costs

Zone 1	2017 carloads	% weight (share of traffic)	Variable cost per car	Weighted zone 1 cost
Railway company ABC	1,600	44.44%	$116.25	$51.66
Railway company XYZ	2,000	55.56%	$125	$69.45
Zone 1 cost per car				$121.11

The Zone 1 variable cost per car in this example is $121.11.

5.0 Contribution to fixed costs

Finally, a system average contribution to fixed costs is added to the variable costs for each zone to arrive at the interswitching rate for the zone. Fixed costs include items that are completely non-variable, such as the maintenance of bridges and snow removal. The costs related to the maintenance of bridges and snow removal do not vary with railway traffic volumes, but are caused by weather and age.

The Agency calculates the system average contribution to fixed costs separately for each railway company. The amount of fixed costs is calculated as the total system cost (which is derived from financial reports provided to the Agency) less the system variable cost (calculated by the Agency’s costing model). The system contribution to fixed costs is the amount of fixed costs expressed as a proportion of the system variable costs.

For 2021, the average contribution to fixed costs is 66.85%, compared to the 2020 value of 59.07% as set out in Determination No. R-2019-230.

The change in the contribution to fixed costs is a function of the difference between the variable unit costs used and the total costs of the railway companies. Agency staff updated the unit costs for CP from 2014 to 2018 and, on the basis of the calculated unit costs, there was an increase in the contribution required according to the methodology for allocating constant costs.

APPENDIX B

Variable costs per shipper:

At year $\left(t0\right)$ for which the last costing information is available:

$V{C}^{t0}(s,i,z,r)={\displaystyle \sum _j}\genfrac{}{}{0.1ex}{}{C_j^{t0}}{y_j\left(r\right)}{v}_j\left(r\right)y_j^{*}(s,i,z,r)$

where:

• s : is a shipper;
• i : is an interchange;
• z : is a zone ;
• r : is a railway company;
• $C_j^{t0}$ is the cost for a specific expense category j for a railway r at time t0;
• $y_j\left(r\right)$ is a system service unit that drives expenses of category j;
• $v_j\left(r\right)$ is the variability factor for the expense category j;
• $y_j^{*}(s,i,z,r)$ is the interswitching service unit that corresponds to category j, it is specific to a shipper (s) that belongs to a specific interchange (i) in a specific zone (z). In addition, the shipper (s) is a client of a railway (r).

To obtain variable costs at the year of the decision (t), inflation factors $(1+p_j)$ are applied to each cost categories j as follows:

VCts,i,z,r=∑jCjt0(r)×(1+pj)yj(r)vj yj*s,i,z,r

 

Inflation factors $(1+p_j)$ that are specific to each expense category j are inserted into the above formula. These inflation factors are developed each year by the Agency.

 

Variable costs per shipper

 

Variable costs per shipper are then averaged over the interchange that they belong to and the railway company that was used. Shippers, interchanges and railway companies are weighted based on their relative share of total carloads.

 

VC(z)=∑r∈z∑i∈z∑s∈iVCt(s,i,z,r) ωs ωiωr

 

Weights:

 

${\chi }_s$ is the volume of cars (measured with carloads) related to a specific shipper (s);

 

ωs=xs∑s∈ixs is the weight of each shipper (s) in a specific interchange (i);

 

ωi=∑s∈ixs ∑i∈z∑s∈ixs is the weight of each interchange (i) in a specific zone (z);

 

ωr=∑i∈z∑s∈ixs ∑r∈z∑i∈z∑s∈ixs is the weight of each railway (r) in a specific zone (z).

 

Section 4.0 (Volume of interswitched cars) of Appendix A provides examples of how the weighted averages are calculated.

 

Final rates per zone:

 

The final rates per zone are obtained by applying a contribution to fixed costs ($Contr$) and a productivity factor ($1+g$) to each variable cost per zone $VC\left(z\right)$

 

Rz=VCz Contr1+g

 

Where the contribution to fixed costs ($Contr$) is estimated using the following equation:

 

Contribution to Fixed Cost (F)= Total system costSystem variable cost

 

the total system cost is obtained from the annual reports of the railway companies, and the system variable cost is calculated by the Agency’s costing model using submitted financial and operating data from the railway companies annually.

 

The productivity variable (g) is provided by:

 

Productivity Index g= Fisher Output IndexFisher Input Index X 100= Ft0,t1OutputFt0,t1Input X 100

 

Fisher Output Index= Ft0,t1Output=Lt0,t1Output×Pt0,t1Output

 

Laspeyres Output Index= Lt0,t1Output= ∑i=1Npi,t0×yi,t1∑i=1Npi,t0×yi,t0=∑i=1Nyi,t1yi,t0×wi,t0

 

Paasche Output Index= Pt0,t1Output= ∑i=1Npi,t1×yi,t1∑i=1Npi,t1×yi,t0=∑i=1Nyi,t1yi,t0×wi,t1

 

Fisher Input Index= Ft0,t1Input=Lt0,t1Input×Pt0,t1Input

 

Laspeyres Input Index= Lt0,t1Input= ∑j=1Mwj,t0×xj,t1∑j=1Mwj,t0×xj,t0=∑j=1Mxj,t1xj,t0×zj,t0

 

Paasche Input Index= Pt0,t1Input= ∑j=1Mwj,t1×xj,t1∑j=1Mwj,t1×xj,t0=∑j=1Mxj,t1xj,t0×zj,t1

 

where:

 

t0 is the base period;

 

t1 is the current period;

 

i is output commodities, and i ranges from 1 to N;

 

j is output commodities, and j ranges from 1 to M;

 

p is the output commodity price;

 

q is the output commodity quantity;

 

w is the input commodity price;

 

x is the input commodity quantity;

 

wi,t0 is the share of ith output commodity in the base period value and wi,t0=pi,t0×yi,t0∑i=1Npi,t0×yi,t0;

 

wi,t1 is the share of ithoutput commodity in the current period value and wi,t1=pi,t1×yi,t1∑i=1Npi,t1×yi,t1

 

zj,t0 is the share of jth input commodity in the base period value and zj,t0=wj,t0×xj,t0∑j=1Mwj,t0×xj,t0;

 

zj,t1 is the share of jth input commodity in the current period value and zj,t1=wj,t1×xj,t1∑j=1Mwj,t1×xj,t1

 

APPENDIX C

 

Accounts from the Uniform Classification of Accounts and related railway records (2014) [UCA] that factor into the 2021 interswitching rates

 

Cost complex	UCA account number	Description of account
102cx	102	Grading
102cx	103	Rail
102cx	105	Ties
102cx	106	Paved Concrete Trackbed (PACT System)
102cx	107	Other Track Material
102cx	109	Ballast
102cx	111	Track Laying and Surfacing
102cx	123	Public Improvements
102cx	125	Other Right-of-Way Property
102cx	139	Roadway Buildings
102cx	141	Roadway Building Machines and Moveable Equipment
131cx	131	Office and Common Buildings
131cx	133	Office and Common Buildings Moveable Equipment and Machinery
143	143	Equipment Repair Shops
145	145	Shop Machinery and Moveable Equipment
149	149	Signals
151	151	Rail Communication Systems
163	163	Fuel Stations
171	171	Locomotives
183	183	Roadway Machines
187cx	187	Work Equipment
187cx	189	Other Non-Revenue Rolling Stock
195	195	Miscellaneous Equipment
400cx	400	Administration
400cx	463	Injuries to Railway Employees: Maintenance of Way and Structures
400cx	479	Other Way and Structure Expense
401cx	401	Track and Roadway Maintenance
401cx	403	Rails – Maintenance
401cx	405	Ties - Maintenance
401cx	406	Paved Concrete Trackbed - Maintenance
401cx	407	Other Track Material - Maintenance
401cx	409	Ballast - Maintenance
401cx	419	Tools and Supplies
401cx	423	Crossing Maintenance
401cx	435	Roadway Buildings - Maintenance
401cx	461	Vehicles
401cx	431	Office and Common Buildings - Maintenance
401cx	437	Equipment Repair Shops - Maintenance
441cx	441	Track Signals - Maintenance
441cx	442	Hump Yard Devices - Maintenance
441cx	443	Crossing Protection - Maintenance
441cx	444	Other Signal Devices - Maintenance
441cx	671	Dispatching
441cx	673	Line Operators and Signal Operation
445cx	445	Rail Communication Systems - Maintenance
445cx	701	Rail Communication System Operation
457	457	Fuel Stations - Maintenance
500cx	500	Administration
500cx	571	Injuries to Railway Employees: Equipment Maintenance
500cx	579	Other Equipment Expense
501	501	Locomotive Maintenance
503	503	Locomotive Servicing
517	517	Lubrication, Inspection and Coupling Hose - Freight Cars
537	537	Work Equipment - Maintenance
539	539	Roadway machines - Maintenance
563cx	563	Work Equipment and Roadway Machine Rents - Dr.
563cx	564	Work Equipment and Roadway Machine Rents - Cr.
573	573	Shop Machinery - Maintenance
600cx	600	Administration
600cx	709	Building Operating Expenses
600cx	711	Other Rail Operations
600cx	743	Injuries to Railway Employees: Rail Operations (Yard and Train)
600cx	745	Clearing Wrecks
600cx	747	Third Party Injuries and Damage to Property (excluding Freight)
600cx	751	Miscellaneous Operating Expense
600cx	607	Train Crews - Freight
619	619	Train Locomotive Diesel Fuel - Freight
631	631	Train Other Expenses - Freight
641cx	641	Controlling Yard Operations
641cx	643	Yard and Terminal Clerical
645cx	645	Yard Engine Crews
645cx	647	Yard Train Crews
645cx	649	Operating Yard Devices
645cx	655	Yard Other Expense
651	651	Yard Locomotive Diesel Fuel
681cx	681	Freight Customer Service Centres
681cx	703	Weighing, Inspection and Demurrage Bureaus
741	741	Loss and Damage: Freight Train Accidents
749	749	Loss and Damage - Other Accidents
800cx	800	General Administration
800cx	801	Management Services
800cx	809	Accounting and Finance
800cx	811	Personnel and Public Relations
800cx	817	Other Administrative Expenses
800cx	861	Injuries to Railway Employees: General (and unallocated)
803	803	Marketing and Sales - Carload Freight
813	813	Environmental Remediation Expense
819	819	Employee Incentive Compensation
821	821	Pension Costs
823cx	823	Health and Welfare
823cx	825	Canada Pension Plan
823cx	827	Quebec Pension Plan
823cx	829	Employment Insurance
831	831	Other Employee Benefits
835	835	Labour Restructuring Expense
843	843	Provincial Sales Taxes
845cx	845	Municipal Property Taxes
845cx	849	Other Taxes
851	851	Insurance
902cx	902	Grading - Amortization
902cx	903	Rail - Amortization
902cx	905	Ties - Amortization
902cx	906	Paved Concrete Trackbed - Amortization
902cx	907	Other Track Material - Amortization
902cx	909	Ballast - Amortization
902cx	911	Track Laying and Surfacing - Amortization
902cx	923	Public Improvements - Amortization
902cx	925	Other Right-of-Way Property - Amortization
902cx	939	Roadway Buildings - Amortization
902cx	941	Roadway Building Machines and Moveable Equipment - Amortization
931cx	931	Office and Common Buildings - Amortization
931cx	933	Office and Common Buildings Moveable Equipment and Machinery - Amortization
943	943	Equipment Repair Shops - Amortization
945	945	Shop Machinery and Moveable Equipment - Amortization
949	949	Signals - Amortization
951	951	Rail Communication Systems - Amortization
963	963	Fuel Stations - Amortization
971	971	Locomotives - Amortization
983	983	Roadway Machines - Amortization
987cx	987	Work Equipment - Amortization
987cx	989	Other Non-Revenue Rolling Stock - Amortization
995	995	Miscellaneous Equipment - Amortization