2019-25924. Waybill Sample Reporting

AGENCY:

Surface Transportation Board.

ACTION:

Notice of proposed rulemaking.

SUMMARY:

The Surface Transportation Board (Board) proposes to amend its regulations with respect to the Waybill Sample data that railroads are required to submit to the Board. The proposed amendments to the Waybill Sample regulations would simplify the sampling rates of non-intermodal carload shipments and specify separate sampling strata and rates for intermodal shipments.

DATES:

Comments are due by January 28, 2020. Replies are due by February 27, 2020.

ADDRESSES:

Comments and replies may be filed with the Board either via e-filing or in writing addressed to: Surface Transportation Board, Attn: Docket No. EP 385 (Sub-No. 8), 395 E Street SW, Washington, DC 20423-0001. Comments and replies will be posted on the Board's website at www.stb.gov.

FOR FURTHER INFORMATION CONTACT:

Jonathon Binet at [phone]. Assistance for the hearing impaired is available through the Federal Relay Service at [phone].

SUPPLEMENTARY INFORMATION:

A waybill is a “document or instrument prepared from the bill of lading contract or shipper's instructions as to the disposition of the freight, and [is] used by the railroad(s) involved as the authority to move the shipment and as the basis for determining the freight charges and interline settlements.” 49 CFR 1244.1(c). Among other things, a waybill currently contains the following data: (1) The originating and terminating freight stations; (2) the railroads participating in the movement; (3) the points of all railroad interchanges; (4) the number and type of cars; (5) the car initial and number; (6) the movement weight in hundredweight; (7) the commodity; and (8) the freight revenue. Rail carriers are required to file a sample of waybills, which includes this data. See 49 CFR 1244.2(a). The Board creates an aggregate compilation of the sampled waybills of all reporting carriers, referred to as the Waybill Sample. The Waybill Sample is the Board's principal source of data about freight rail shipments. It has broad application in, among other things, rate cases, the development of costing systems, productivity studies, exemption decisions, and analyses of industry trends.

First collected in 1946 by the Board's predecessor,^[1] the Interstate Commerce Commission (ICC), the Waybill Sample is also used by other Federal agencies, state and local government agencies, the transportation industry, shippers, research organizations, universities, and others that have a need for rail shipment data. Because some of the submitted waybill data is commercially sensitive, the Board's regulations place limitations on the release and use of confidential Waybill Sample data. See 49 CFR 1244.9; see also 49 U.S.C. 11904.^[2]

In January 2018, the Board established its Rate Reform Task Force (RRTF), with the objectives of developing recommendations to reform and streamline the Board's rate review processes for large cases, and determining how to best provide a rate review process for smaller cases. After holding informal meetings throughout 2018, the RRTF issued a report on April 25, 2019 (RRTF Report).^[3] Among other recommendations, the RRTF Report included a recommendation that the Board change the sampling rates for its Waybill Sample. RRTF Report 14. The RRTF explained that data from the Waybill Sample is critical to certain rate cases, in particular the Three-Benchmark methodology, and that a more robust sample size would address issues with those cases. Id. at 47. Having considered the recommendations included in the RRTF Report and the overall utility of the current Waybill Sample, the Board now proposes to simplify the sampling rate for non-intermodal carload shipments and specify separate sampling strata and rates for intermodal shipments, as explained below.

Current Waybill Sampling Requirements. A railroad is required to file with the Board a sample of its waybill data for all line-haul revenue waybills terminated on its lines in the United States,^[4] if the railroad: (a) Terminated at least 4,500 revenue carloads in any of the three preceding years, or (b) terminated at least 5% of the revenue carloads terminating in any state in any of the three preceding years. 49 CFR 1244.2(a). Currently, the number of waybills that a railroad is required to file (i.e., the sampling rate) is set forth at 49 CFR 1244.4(b) and (c), and varies based on the number of carloads on the waybill.^[5] The current sampling rates for the computerized system of reporting waybills are shown in Table 1 below.

Table 1—Current Waybill Sampling Rates

[Computerized System of Reporting]

Number of carloads on waybill	Sample rate 6
1 to 2	1/40
3 to 15	1/12
16 to 60	1/4
61 to 100	1/3
101 and over	1/2

RRTF Proposal and Board Rationale. In its report, the RRTF recommended changing the current waybill sampling rates for all non-intermodal shipments to 1/10. RRTF Report 48. For intermodal shipments, the RRTF recommended two strata: (1) Intermodal shipments with one or two trailer or container units (TCUs) per waybill, recommended to be sampled at the current 1/40 rate, and (2) intermodal shipments with three or more TCUs per waybill, recommended to be sampled at the same rate proposed for non-intermodal shipments, 1/10. Id. at 48-49. Although these recommendations would both increase the sampling rates for most smaller shipments (with 1 to 15 carloads per waybill) from 1/40 or 1/12 to 1/10 and decrease the sampling rates for larger shipments (with 16 or more carloads per waybill) from 1/2, 1/3, or 1/4 to 1/10, the RRTF determined that the net effect of the recommended changes would be an increase in the overall number of waybills sampled. Id. at 48. In addition, by sampling intermodal traffic separately and (for one or two TCUs) at the current 1/40 rate, the RRTF concluded that a greater portion of the Waybill Sample data would represent regulated traffic instead of traffic that is currently exempt.^[7] Id. at 49.

The RRTF supported its recommendation by describing the anticipated effect the changes would have in rate cases under the Board's Three-Benchmark methodology.^[8] The RRTF stated that, by increasing its sampling of traffic, “the Board could avoid the scarcity issue that has plagued some past Three-Benchmark cases.” Id. at 47. See, e.g., US Magnesium, L.L.C. v. Union Pac. R.R., NOR 42114, slip op. at 9 n.12 (STB served Jan. 28, 2010) (“We acknowledge that the failure of either party to submit a comparison group more similar to the traffic at issue here is likely due to limitations in the number of comparable movements in the Waybill Sample.”). The RRTF also stated that a robust sample size is a critical component of the Three-Benchmark methodology and explained that there must be enough observations in the Waybill Sample to select a group of traffic that reflects the nuances of the traffic in dispute. RRTF Report 47. It stated that its recommendation to modify waybill sampling rates would alleviate concerns about non-representative samples and minimize the need for “other relevant factors” arguments. Id.

The Board agrees with the RRTF that a modification to its waybill sampling rates is warranted. Specifically, a net increase in sample size would provide more comprehensive information to the Board and other users of Waybill Sample data in a variety of contexts, such as exemption decisions, stratification reports, traffic volume and rate studies, Board-initiated investigations, certain rate cases (discussed in more detail below), and any other waybill data-related analysis the Board currently performs or might seek to perform in the future. A more robust data sample would augment the Board's ability to make informed, well-reasoned decisions in these areas. In addition, the Board agrees that it should change its sampling requirements so that a greater portion of the Waybill Sample data would represent regulated traffic instead of exempt traffic.

Additionally, the added number of observations in the Waybill Sample would likely allow the Board to avoid redacting, for confidentiality reasons, as many results from some of the Board's routine analysis published on its website, such as the STCC 7 stratification report. While such analysis serves as a useful barometer for stakeholders, its publication is limited by the Board's commitment to protect the confidentiality of identifiable railroad and shipper information when too few records exist within a given category of traffic. Moreover, because the Board currently receives monthly waybill data from Class I carriers and quarterly data from Class I, II, and III carriers, increasing the sampling rate would provide the Board with more observations in any given month or quarter from which it could draw meaningful insights throughout the year.

The Board agrees with the RRTF that increasing waybill sampling rates would also assist parties in Three-Benchmark cases by providing a greater number of potentially comparable movements from which they could create their comparison group proposals.^[9] Parties proposing comparison groups use a variety of comparability factors, such as the length of movement, commodity type, and traffic densities of the likely routes involved. In general, as more comparability factors are added to make the comparison group more specific to the case, the number of observations from the Waybill Sample that match those factors is likely to decrease. By increasing the observations in the Waybill Sample, parties generally would have more observations to choose from and increased flexibility to design comparison groups with relevant comparability factors. Accordingly, a more robust Waybill Sample could lead to more representative comparison groups, thereby increasing the reliability of the parties' presentations.

The increased comparison group flexibility would also increase the number of potentially comparable movements available to shippers of categories of traffic for which there are currently insufficient observations in the Waybill Sample to create a representative comparison group.^[10] The Board's proposed changes in sampling, discussed below, would result in more shipments being included in the Waybill Sample, some of which may fall into categories of traffic that previously had fewer than 25 movements in the Waybill Sample.^[11] Moreover, for the reasons noted above, even for categories of traffic for which a comparison group with 25 or more observations can already be formed, more observations in the Waybill Sample could allow for the addition of more specific traffic characteristics and would further increase the reliability of the parties' presentations.^[12]

The issue of whether to enlarge the Waybill Sample to include a larger sample of common carrier movements was briefly discussed in the Board's decision in Simplified Standards, EP 646 (Sub-No. 1), slip op. at 83. There, the Board declined to increase the waybill sampling size at that time due to concerns about the cost of gathering, processing, and costing a larger sample. However, it is now appropriate to revisit the issue. The Board finds that the expenses associated with increased sampling can be better managed by the agency because of technological and computing advances now available to it, and finds that the largely computerized and automated processes allow for the management of additional data at a reasonable additional cost. All reporting carriers submit waybill data in computerized form today, and the Board does not anticipate that it would be a significant burden for rail carriers, or the entity the carriers use to manage the data, to adjust their data collection and reporting mechanism(s) for the proposed sampling rates. Given that this data is critical to central regulatory functions of the Board, the additional cost is justified by the anticipated improvements in reliability of comparison group presentations and by the increased granularity of analyses performed by the Board.

For the reasons discussed above, the Board proposes to adjust the waybill sampling rate for carriers using the computerized system of reporting as discussed below. The Board's proposal is intended to provide a more comprehensive sampling of waybills that would improve the utility of the Waybill Sample for both the Board and other users of waybill data in a variety of contexts (e.g., increasing the reliability of parties' evidentiary presentations in certain rate reasonableness proceedings), which would further the rail transportation policy goals of 49 U.S.C. 10101. See 49 U.S.C. 10101(2), (4), (6), (13).

Proposed Waybill Sampling Rates. The Board proposes revisions to the sampling rates for the Waybill Sample for carriers using the computerized system of reporting. Although the RRTF recommended a sampling rate for all non-intermodal shipments of 1/10, based on additional analyses, described below, the Board instead proposes to increase the sampling rates to 1/5 for non-intermodal shipments in each of the existing sampling strata, as shown in Table 2 below. Under this proposed rule, the Board would continue to use separate strata for the sampling of non-intermodal shipments, with the strata differentiated by the number of carloads on the waybill. For non-intermodal shipments, the effect of the proposed rate would be an increase in the sampling rate for waybills with 1 to 15 carloads and a decrease in the sampling rate for waybills with 16 or more carloads.

Because of the unique characteristics of intermodal shipment billing practices,^[13] the Board also proposes to separate sampling of intermodal shipments from carload shipments. Specifically, the Board would create two sampling strata specific to intermodal shipments—one for shipments with one to two TCUs per waybill and another for shipments with three or more TCUs per waybill. As shown in Table 2, intermodal shipments with one or two TCUs per waybill would be sampled at a rate of 1/40, and intermodal shipments with three or more TCUs per waybill would be sampled at the same proposed rate as non-intermodal shipments, 1/5. An increase in sampling of intermodal shipments with one or two TCUs per waybill, which comprise the vast majority of intermodal shipments, would lead to an over-sampling of those movements.^[14] The Board's proposed approach would not only appropriately differentiate sampling strata based on industry waybill practices, but it would also avoid instances in which blocks of TCUs comprising a single intermodal shipment are over-sampled.

The Board's proposal for intermodal shipments largely mirrors the RRTF recommendation that the Board adopt a sampling rate of 1/40 for waybills with one to two TCUs and apply the same sampling rate recommended for non-intermodal shipments for waybills with three or more TCUs. Consistent with the approach recommended by the RRTF, the Board proposes the same sampling rate for intermodal waybills with three or more TCUs as it proposes for non-intermodal shipments.

Table 2—Proposed Waybill Sampling Rates

[Computerized System of Reporting) ¹⁵

Number of non-intermodal carloads on waybill	Sample rate
1 to 2	1/5
3 to 15	1/5
16 to 60	1/5
61 to 100	1/5
101 and over	1/5
Number of intermodal trailer/container units on waybill	Sample rate
1 to 2	1/40
3 and over	1/5

Analysis of Proposed Waybill Sampling Rates. As discussed above, these proposed changes would both provide a more robust sample generally and address the shortcomings that were acknowledged by the Board and parties in Board proceedings concerning the scarcity of data in some rate cases. See US Magnesium, L.L.C., NOR 42114, slip op. at 9-12, 9 n.12 (noting the dearth of observations for certain toxic-by-inhalation commodities in the parties' comparison groups); Simplified Standards, EP 646 (Sub-No. 1), slip op. at 83 (acknowledging that there may be instances in Three-Benchmark cases where a particular movement is so unique that there would be insufficient comparable movements in the Waybill Sample).

To determine the impact of increasing its sampling rates, the Board has reviewed Waybill Sample data from 2014 to 2017; grouped the movements into categories based on commodity,^[16] mileage ranges,^[17] and terminating railroad; and analyzed how the proposed sampling rates would affect the number of these movement categories having fewer than 25 observations.^[18] Under the current sampling rates, the Board found that, in an average year, approximately 7.6% of those movement categories have 25 or more observations. Under the Board's proposed sampling rates, an estimated 20.4% of those categories would have 25 or more observations, nearly triple the current number. Even though only one-fifth of the categories would have at least 25 observations under the Board's proposal, this segment represents most of the total revenue in the Waybill Sample. Under the current sampling rate, 84.2% of the revenue is represented in movement categories with at least 25 observations. Under the proposed sampling rate, 93.4% of the revenue would be represented in movement categories with at least 25 observations. The proposed modification would therefore capture more than half of the revenue that is currently moving in categories with fewer than 25 movements. These percentage breakdowns are shown in Table 3 below.

Table 3—Estimated Movement Categories in an Average Year

[2014-2017]

	Total movement categories	Movement categories with 25+ observations	Percent of movement categories with 25+ observations	Percent of revenue in movement categories with 25+ observations
Current Rate	31,321	2,369	7.6	84.2
Proposed Rate	31,321	6,395	20.4	93.4

Similarly, under the current regulations, when aggregating the Waybill Sample data over the four-year 2014 to 2017 period, 19.7% of the same categories include 25 or more observations. Using the four-year approach, under the proposed sampling rate, the number of categories with 25 or more observations would nearly double to 38.5%. Here the proposed modification would capture approximately two-thirds of the currently missed revenue, increasing from 94.0% to 97.9% of total revenue. This breakdown is shown in Table 4 below.

Table 4—Estimated Movement Categories Over Four Years

[2014-2017]

	Total movement categories	Movement categories with 25+ observations	Percent of movement categories with 25+ observations	Percent of revenue in movement categories with 25+ observations
Current Rate	31,321	6,177	19.7	94.0
Proposed Rate	31,321	12,059	38.5	97.9

The Board considered the 1/10 sampling rate for non-intermodal shipments recommended by the RRTF. The Board's analysis, however, showed that a 1/5 sampling rate had a better chance of reducing the number of movement categories with scarce observations. Although the improvement was modest—for example, 90.4% of the revenue in the Waybill Sample would be in movement categories with 25 or more observations in an average year with a 1/10 sampling rate compared to 93.4% with a 1/5 sampling rate—the potential increase in covered movement categories would lead to a more robust sample without a significantly increased burden on reporting carriers. While the waybill sampling rates listed in Table 2 above may still, in some instances, fail to produce a representative sample for comparison, the proposed changes would significantly improve the chances of having sufficient observations for a representative sample as well as add to the robustness of any of Board analyses using the Waybill Sample.

For movement categories that already have 25 or more observations, such as traffic in categories with a higher volume of movements by rail, the Board analyzed the extent to which more observations in the Waybill Sample would allow for more granular or even additional comparability factors. As can be seen in Table 5 below, the Board estimates that the proposed sampling rate would increase the median number of observations for categories that already have at least 25 observations in an average year from 59 to 269, which is more than four times as many observations. This illustrates how the proposed sampling rate would shift the number of observations upwards across categories, even if the categories already had 25 observations. Such an increase in observations would increase the representativeness of potential comparison groups defined using the same criteria as these categories. Furthermore, as noted above, by having more observations in a comparison group, it would be possible to define the comparison group even more narrowly and still maintain robustness.

Table 5—Quartile Analysis of Movement Categories With 25+ Observations in an Average Year

[2014-2017]

	1st Quartile observations	Median observations	3rd Quartile observations
Current Rate	37	59	126
Proposed Rate	101	269	562

Table 6 below shows similar estimated increases in observations over four years of data. The Board estimates that the proposed sampling rate would increase the median number of observations for categories that already have at least 25 observations over the course of a four-year period from 70 to 320.

Table 6—Quartile Analysis of Movement Categories With 25+ Observations Over Four Years

[2014-2017]

	1st Quartile observations	Median observations	3rd Quartile observations
Current Rate	39	70	177
Proposed Rate	136	320	780

Once again, the proposed sampling rate is estimated to result in more than four times as many observations as under the current rate. For example, consider the median category with at least 25 observations over four years as shown in Table 6. Using a 500-mile range as a comparability factor, a party would have 70 observations to include in a potential comparison group. If that party wanted to define the mileage range more narrowly, they would lose some of those observations depending on the mileage range chosen and, at some point, would likely have fewer than 25 observations. If, however, a party started with 320 available observations with a 500-mile range, as we estimate would be the case in the median category under the proposed sampling rates, they could likely narrow the mileage range further without dropping below a sufficient number of observations. In other words, with more observations available, interested parties would be able to choose additional and more narrow comparability factors to identify movements that are more similar to the issue traffic but also still maintain a sufficient number of observations.

Conclusion. For the reasons described above, the changes proposed in this NPRM (as shown below) would create a more robust Waybill Sample and result in more comprehensive information that would assist both the Board in its decision-making and analyses and other users of waybill data in their analyses without creating an undue burden on railroads (as shown below and in the Appendix). The changes also appropriately differentiate sampling strata based on current industry waybill practices for intermodal shipments. The Board invites public comment on this proposal.

Regulatory Flexibility Act. The Regulatory Flexibility Act of 1980 (RFA), 5 U.S.C. 601-612, generally requires a description and analysis of new rules that would have a significant economic impact on a substantial number of small entities. In drafting a rule, an agency is required to: (1) Assess the effect that its regulation would have on small entities; (2) analyze effective alternatives that may minimize a regulation's impact; and (3) make the analysis available for public comment. Section 601-604. In its notice of proposed rulemaking, the agency must either include an initial regulatory flexibility analysis, section 603(a), or certify that the proposed rule would not have a “significant impact on a substantial number of small entities,” section 605(b). Because the goal of the RFA is to reduce the cost to small entities of complying with federal regulations, the RFA requires an agency to perform a regulatory flexibility analysis of small entity impacts only when a rule directly regulates those entities. In other words, the impact must be a direct impact on small entities “whose conduct is circumscribed or mandated” by the proposed rule. White Eagle Coop. v. Conner, 553 F.3d 467, 480 (7th Cir. 2009). An agency has no obligation to conduct a small entity impact analysis of effects on entities that it does not regulate. United Dist. Cos. v. FERC, 88 F.3d 1105, 1170 (DC Cir. 1996).

This proposal would not have a significant economic impact upon a substantial number of small entities, within the meaning of the RFA.^[19] Under the Board's existing regulations, a railroad is required to file Waybill Sample data for all line-haul revenue waybills terminated on its lines if: (a) It terminated at least 4,500 revenue carloads in any of the three preceding years; or (b) it terminated at least 5% of the revenue carloads terminating in any state in any of the three preceding years. 49 CFR 1244.2. Under this criteria, 53 railroads are currently required to report Waybill Sample data. Of these 53, the Board estimates that 36 are Class III carriers, and thus small businesses within the meaning of the RFA. Of the 53 railroads required to report Waybill Sample data, 45 railroads currently use Railinc Corporation (Railinc)—a wholly-owned information technology subsidiary of the Association of American Railroads—to sample their waybills.^[20] Eight railroads currently sample their own waybills.

For the railroads that submit their waybills to Railinc for sampling, there would be no additional burden or costs on entities as result of the changes proposed in this NPRM. These entities would continue to submit all of their waybills to Railinc, which would then sample the data in accordance with the Board's revised sampling rates. Because the Board contracts with Railinc to sample railroads' waybills, the entities that use Railinc to sample their waybills would incur no additional costs from Railinc as a result of the Board's proposed changes. Of the approximately 36 Class III carriers, the Board estimates that 34 fall into this category and therefore would not incur any additional burden or cost.

For the railroads that choose to sample their own waybills, the proposed amendments would not result in a significant economic impact. The purpose of the changes proposed in this NPRM is to create a more robust Waybill Sample, resulting in more comprehensive information that would assist both the Board in its decision-making and analyses and other users of waybill data in their analyses. The proposal would increase the rate at which the Board samples certain railroad shipments and appropriately differentiate sampling strata based on industry waybill practices for intermodal shipments. These changes would result in additional observations for certain shipments, but the proposed amendments would not significantly alter small entities' current practices for sampling their shipments. Based on the total burden hours described in the Paperwork Reduction Act analysis below, the Board estimates that, for railroads conducting their own sampling, the change in reporting procedures would result in an estimated one-time burden of approximately 80 hours per railroad. Moreover, this impact would not be on a substantial number of small entities, as the Board estimates that only two of the approximately 36 Class III carriers would incur this burden.

For the reasons described above, the Board certifies under 5 U.S.C. 605(b) that this proposed rule, if promulgated, would not have a significant economic impact on a substantial number of small entities within the meaning of the RFA.

Paperwork Reduction Act. Pursuant to the Paperwork Reduction Act (PRA), 44 U.S.C. 3501-3521, Office of Management and Budget (OMB) regulations at 5 CFR 1320.8(d)(3), and in the Appendix, the Board seeks comments about the impact of the revisions in the proposed rules to the currently approved collection of Waybill Sample data (OMB Control No. 2140-0015) regarding: (1) Whether the collection of data, as modified in the proposed rule and further described below, is necessary for the proper performance of the functions of the Board, including whether the collection has practical utility; (2) the accuracy of the Board's burden estimates; (3) ways to enhance the quality, utility, and clarity of the data collected; and (4) ways to minimize the burden of the collection of data on the respondents, including the use of automated collection techniques or other forms of information technology, when appropriate.

The Board estimates that the proposed requirements would add a total one-time hour burden of 640 hours (or approximately 213.3 hours per year as amortized over three years) because the railroads, in most cases, would need to edit their software programs to implement these changes. Once the burden of the one-time programming changes is incurred, the annual burden would remain the same as before this modification. The Board welcomes comment on the estimates of actual time and costs of collection of Waybill Sample data, as detailed below in the Appendix.^[21] The proposed rules will be submitted to OMB for review as required under 44 U.S.C. 3507(d) and 5 CFR 1320.11. Comments received by the Board regarding the data collection will also be forwarded to OMB for its review when the final rule is published.

List of subjects in 49 CFR Part 1244

• Freight
• Railroads
• Reporting and recordkeeping requirements

It is ordered:

1. The Board proposes to amend its rules as detailed in this decision. Notice of the proposed rules will be published in the Federal Register.

2. Comments are due by January 28, 2020. Replies are due by February 27, 2020.

3. A copy of this decision will be served upon the Chief Counsel for Advocacy, Office of Advocacy, U.S. Small Business Administration, 409 3rd Street SW, Washington, DC 20416.

4. This decision is effective on its service date.

Decided: November 22, 2019.

By the Board, Board Members Begeman, Fuchs, and Oberman.

Kenyatta Clay,

Clearance Clerk.

For the reasons set forth in the preamble, the Surface Transportation Board proposes to amend part 1244 of title 49, chapter X, of the Code of Federal Regulations as follows:

PART 1244—WAYBILL ANALYSIS OF TRANSPORTATION OF PROPERTY—RAILROADS

1. The authority citation for part 1244 continues to read as follows:

Authority: 49 U.S.C. 1321, 10707, 11144, 11145.

2. Amend § 1244.4 by revising the first sentence of paragraph (c)(1) and replacing the current table in paragraph (c)(2) with a new table to read as follows:

§ 1244.4

Sampling of waybills.

* * * * *

(c) The Computerized System. (1) The tape shall be required to conform to the standards and format specified in Statement No. 81-1, Procedure for Sampling Waybill Records by Computer (2019 edition), issued by the Surface Transportation Board.

* * * * *

(2) Effective January 1, 2021, and thereafter, unless otherwise ordered, the sampling rates for the computerized system are as follows:

Number of non-intermodal carloads on waybill	Sample rate
1 to 2	1/5
3 to 15	1/5
16 to 60	1/5
61 to 100	1/5
101 and over	1/5
Number of intermodal trailer/container units on waybill	Sample rate
1 to 2	1/40
3 and over	1/5

* * * * *

Note: This appendix will not appear in the Code of Federal Regulations

Appendix

Information Collected Under the Paperwork Reduction Act

Title: Waybill Sample.

OMB Control Number: 2140-0015.

Form Number: None.

Type of Review: Revision of a currently approved collection.

Summary: As part of its continuing effort to reduce paperwork burdens, and as required by the Paperwork Reduction Act of 1995 (PRA), 44 U.S.C. 3501-3521, the Surface Transportation Board (Board) gives notice that it is requesting from the Office of Management and Budget (OMB) approval for the revision of the currently approved data collection, Waybill Sample, OMB Control No. 2140-0015, as further described below. The requested revision to the currently approved collection is necessitated by this Notice of Proposed Rulemaking (NPRM), which would amend the Waybill Sample data railroads are required to submit to the Board pursuant to 49 CFR 1244.4. All other data collected by the Board in the currently approved collection is without change from its approval (currently expiring on September 30, 2020).

Respondents: Respondents include any railroad that is subject to the Interstate Commerce Act and that terminated at least 4,500 carloads on its line in any of the three preceding years or that terminated at least 5% of the revenue carloads terminating in any state in any of the three preceding years. For the purposes of this analysis, the Board categorizes railroads required to report Waybill Sample data as either quarterly or monthly and as either sampling their own waybills or having a third party conduct their sampling. As a result, there are four categories of respondents, as shown in Table B-1 below.

Table B-1—Respondents

Categories of respondents	Number of respondents
Railroads that conduct their own sampling and report monthly	5
Railroads that conduct their own sampling and report quarterly	3
Railroads that have a third party sample their waybills and report monthly	2
Railroads that have a third party sample their waybills and report quarterly	43

Number of Respondents: 53.

Estimated Time Per Response: The estimated hour burden for waybill samples submitted to the Board is shown in Table B-2 below. (Note: respondents that are identified as reporting monthly actually report monthly, quarterly, and annually (or 17 times per year). All other respondents report quarterly and annually (five times a year)). The annualized one-time hour burden resulting from this NPRM is shown in Table B-3 below.

Table B-2—Estimated Existing Annual Hour Burden Under Current Regulations

Categories of respondents	Number of respondents	Total number of samples submitted	Estimated annual hours per sample submitted	Total estimated annual hours for samples submitted
Railroads that conduct their own sampling and report monthly	5	85	2.5	212.5
Railroads that conduct their own sampling and report quarterly	3	15	2.5	37.5
Railroads that have a third party sample their waybills and report monthly	2	34	1.25	42.5
Railroads that have a third party sample their waybills and report quarterly	43	215	1.25	268.8
Total Annual Hour Burden				561.3

Table B-3—Estimated Additional One-Time Hour Burden Under Proposed Regulations

Categories of respondents	Number of respondents	Estimated annual one-time hour burden (per respondent)	Total annual one-time hour burden
Railroads that conduct their own sampling and report monthly	5	26.7	133.3
Railroads that conduct their own sampling and report quarterly	3	26.7	80.0
Railroads that have a third party sample their waybills and report monthly	2	* 0	* 0
Railroads that have a third party sample their waybills and report quarterly	43	* 0	* 0
Total Annual One-Time Hour Burden			213.3
* The Board pays for the third-party contractor to prepare samples. There is no one-time hourly or non-hourly burden to these railroads.

Frequency of Response: Seven respondents report monthly; 46 report quarterly.

Total Burden Hours (annually including all respondents): 774.6 hours. This estimated total burden hours is shown in Table B-4 below.

Table B-4—Total Burden Hours

Categories of respondents	Estimated annual hours for samples submitted	Estimated annual one-time hour burden (amortized over 3 years)	Total annual hour burden
Railroads that conduct their own sampling and report monthly	212.5	133.3	345.8
Railroads that conduct their own sampling and report quarterly	37.5	80.0	117.5
Railroads that have a third party sample their waybills and report monthly	42.5	* 0	42.5
Railroads that have a third party sample their waybills and report quarterly	268.8	* 0	268.8
Total Annual Burden Hours	561.3	213.3	774.6
* The Board pays for the third-party contractor to prepare samples. There is no one-time hourly or non-hourly burden to these railroads.

Total Annual “Non-Hour Burden” Cost: There are no other costs identified because filings are submitted electronically to the Board.

Needs and Uses: The Board is, by statute, responsible for the economic regulation of common carrier rail transportation in the United States. The information in the Waybill Sample is used by the Board, other federal and state agencies, and industry stakeholders to monitor traffic flows and rate trends in the industry, and to develop testimony in Board proceedings. The Board has authority to collect this data under 49 U.S.C. 11144 and 11145. As described in more detail above in the NPRM, the Board is amending the rules that apply to the collection of the Waybill Sample to simplify the sampling rates of non-intermodal carload shipments and to create more accurate sampling strata and rates for intermodal traffic. The Board's collection and use of this data enables the agency to meet its statutory duty to regulate the rail industry.

Footnotes

1.  See Bureau of Transp. Econ. & Stat., Interstate Com. Comm'n, Statement No. 543, Waybill Statistics their History and Uses 15, 19, 40 (1954); Waybill Analysis of Transp. of Prop.—R.Rs., 364 I.C.C. 928, 929 (1981) (“Since 1946, the Interstate Commerce Commission has collected a continuous sample of carload waybills for railroads terminating shipments.”).

2.  Any grant of access to confidential Waybill Sample data requires the requestor to execute a confidentiality agreement before receiving the data. See 49 CFR 1244.9(a)-(e). In addition to the confidential Waybill Sample, the Board also generates a Public Use Waybill File that includes only non-confidential data. See 49 CFR 1244.9(b)(5).

3.  The RRTF Report was posted on the Board's website on April 29, 2019, and can be accessed at https://www.stb.gov/​stb/​rail/​Rate_​Reform_​Task_​Force_​Report.pdf.

4.  A railroad moving traffic on the United States rail system to the Canadian or Mexican border is required to “include a representative sample of such international export traffic in the Waybill Sample.” 49 CFR 1244.3(c).

5.  The Board's regulations set forth different sampling rates for computerized and manual systems of reporting. See 49 CFR 1244.4(b)-(c). Under the manual system, railroads submit Waybill Sample data through authenticated copies of a sample of audited revenue waybills instead of using a computerized system. Id. section 1244.4(a). The manual system is not currently used by any railroads and is not the primary subject of this notice of proposed rulemaking (NPRM). However, parties may provide comments on whether the manual system should be eliminated given its current lack of use.

6.  The column showing the sample rate indicates the fraction of the total number of waybills within each stratum that must be submitted (e.g., for waybills of one to two carloads, the railroad must submit one out of every 40 waybills).

7.  Under 49 CFR 1090.2, rail and highway trailer-on-flatcar/container-on-flatcar (TOFC/COFC) service—which generally covers intermodal shipments—is exempt from the requirements of 49 U.S.C. subtitle IV, regardless of the type, affiliation, or ownership of the carrier performing the highway portion of the service.

8.  The Three-Benchmark methodology is a simplified process of rate review, intended for smaller rate disputes, where the potential rate relief is capped at $4 million. See Simplified Standards for Rail Rate Cases (Simplified Standards), EP 646 (Sub-No. 1) (STB served Sept. 5, 2007), aff'd in part sub nom. CSX Transp., Inc. v. STB, 568 F.3d 236 (D.C. Cir. 2009), vacated in part on reh'g, 584 F.3d 1076 (D.C. Cir. 2009); Rate Regulation Reforms, EP 715 (STB served July 18, 2013), remanded in part sub nom. CSX Transp., Inc. v. STB, 754 F.3d 1056 (D.C. Cir. 2014). Under this methodology, the reasonableness of a challenged rate is judged by examining the challenged rate using three benchmark figures, each of which is expressed as a revenue-to-variable cost (R/VC) ratio. One of the benchmarks, R/VC_COMP, requires selection of a group of comparable traffic, the “comparison group,” that the Board concludes is most similar in aggregate to the issue movements. To “enable a prompt, expedited resolution of the comparison group selection,” the Board requires each party to submit its final offer comparison group simultaneously, and the Board chooses one of those groups without modification. See Simplified Standards, EP 646 (Sub-No. 1), slip op. at 18.

9.  In a recently issued NPRM, the Board proposed a new procedure for challenging the reasonableness of railroad rates in smaller cases. See Final Offer Rate Review, EP 755 et al. (STB served Sept. 12, 2019). In that decision, the Board stated that, under the proposed Final Offer Rate Review (FORR) procedure, a party would be able to seek access to waybill data pursuant to the Board's regulations. Id. at 9. The benefits of increased waybill sampling discussed in this NPRM could also apply to the proposed FORR procedure, should a party choose to use comparable traffic to support its final offer.

10.  According to the Central Limit Theorem, once a sample has sufficient observations, it is considered to be normally distributed and can be used to approximate the mean and variance of the population from which it was sampled. Generally, around 25 or 30 observations is considered to be enough for those approximations. See Robert V. Hogg et al., Probability and Statistical Inference 202 (9th ed. 2015). In Rail Transp. of Grain, Rate Reg. Review, EP 665 (Sub-No. 1) et al., slip op. at 13-14 (STB served Aug. 31, 2016), the Board expressed concern about comparison groups with insufficient observations and sought comment on whether a 20-observation minimum should be established in connection with a new comparison group approach it was exploring in that proceeding. Because the Board seeks to improve significantly the utility of the Waybill Sample in this proposal, it has used a 25-observation minimum for the purposes of analyzing this proposed rule.

11.  Based on an analysis of the 2014 through 2017 Waybill Samples, this tends to be the case for groups of traffic that do not have as high of a volume of movements as others, meaning that fewer of those movements are captured in the Waybill Sample.

12.  For example, comparability factors such as length of movement ranges could be tightened and more granular commodity codes could be used (e.g., seven-digit STCC level versus five-digit STCC level). In some cases, geographic comparability could be taken into consideration to make the comparison group more similar to the traffic at issue. Currently, depending on the commodity group at issue, the application of such specific criteria could result in a comparison group without sufficient observations.

13.  In a separate proceeding that has since been discontinued, commenters noted that intermodal TCUs often move under separate waybills, even if the TCUs are placed on flatcars that move in multiple flatcar blocks. See Review of the Gen. Purpose Costing Sys., EP 431 (Sub-No. 4), slip op. at 13 (STB served Aug. 4, 2016).

14.  To illustrate, under the Board's current regulations, a block carrying 100 TCUs, all moving from the same origin to the same destination but with each moving under a separate waybill (i.e., 100 total waybills), would be sampled at an average of 2.5 times (i.e., 100 waybills sampled at a rate of 1/40). Under the Board's proposed regulations, if intermodal shipments were sampled at the same rate as non-intermodal shipments, the same large block would ultimately be sampled 20 times (i.e., 100 waybills sampled at a rate of 1/5). Considering this billing practice, along with the volume of intermodal shipments and the fact that intermodal transportation is generally exempt from Board regulation, the Board finds increasing the sampling rate of intermodal shipments with one to two TCUs per waybill is not necessary. By establishing separate sampling strata for intermodal shipments as proposed, the Board can avoid over-sampling intermodal traffic with one or two TCUs per waybill by maintaining the current rate of 1/40, in which case the same large block carrying 100 TCUs would be sampled 2.5 times, as it would be under the current regulations.

15.  If the Board ultimately adopts changes to 49 CFR part 1244, the Board will publish notice in the Federal Register of a revised edition of Statement No. 81-1, Procedure for Sampling Waybill Records by Computer (2009 edition). See 49 CFR 1244.4(c)(1) (requiring the Board to publish notice of any change to Statement No. 81-1 in the Federal Register). The current edition of Statement No. 81-1 is posted on the Board's website and can be accessed by navigating to the tab Industry Data, the tab Economic Data, and then clicking on the link for “Procedure for Sampling Waybill Records by Computer.”

16.  Commodity categories were split at the seven-digit STCC level.

17.  Mileage ranges were split as follows: 0-499.9 miles; 500-999.9 miles; 1,000-1,499.9 miles; and 1,500 miles or more.

18.  In order to estimate how counts of observations would change with the proposed sampling rate, the Board took the observations currently in the Waybill Sample, extrapolated how many observations exist in the total population of movements that occurred in a given year by multiplying counts of movements by their expansion factors, and then divided by five for non-intermodal movements and by 40 for intermodal movements. This is a slight simplification of the Board's proposed sampling rates, since it does not distinguish intermodal movement sampling rates depending on the number of TCUs, but it is reasonable for analysis purposes because the vast majority of intermodal moves are under the three TCU threshold.

19.  For the purpose of RFA analysis for rail carriers subject to Board jurisdiction, the Board defines a “small business” as only including those rail carriers classified as Class III carriers under 49 CFR 1201.1-1. See Small Entity Size Standards Under the Regulatory Flexibility Act, EP 719 (STB served June 30, 2016) (with Board Member Begeman dissenting). Class III carriers have annual operating revenues of $20 million or less in 1991 dollars, or $39,194,876 or less when adjusted for inflation using 2018 data. Class II carriers have annual operating revenues of less than $250 million or $489,935,956 when adjusted for inflation using 2018 data. The Board calculates the revenue deflator factor annually and publishes the railroad revenue thresholds in decisions and on its website. 49 CFR 1201.1-1; Indexing the Annual Operating Revenues of R.Rs., EP 748 (STB served June 14, 2019).

20.  Some railroads hire a third party to collect their waybills. That third party then sends these waybills to Railinc for sampling.

21.  In the Appendix, Tables B-2, B-3, and B-4 show a total annual burden of 774.6 hours, incorporating the annualized one-time hour burden of 213.3 hours under the proposed rule, and the existing annual burden of 561.3 hours.