457 F.3d 31 | D.C. Cir. | 2006
EchoStar Satellite L.L.C., a provider of direct-to-the-home satellite television service, petitions for review of two orders in which the Federal Communications Commission adopted an improved version of its Individual Location Longley-Rice (ILLR) model for predicting the strength of broadcast television signals. In response to con
I. Background
A network television broadcaster generally has “exclusive rights,” pursuant to the Copyright Act, 17 U.S.C. § 106(5), to authorize the public display of its copyrighted content, including the retransmission of its broadcast signal. In the Satellite Home Viewer Act of 1998 (SHVA), the Congress, in order to make network programming available to households outside the broadcast range of a local network affiliate, gave satellite carriers a “statutory license,” that is, a compulsory license subject to payment of a prescribed royalty, to retransmit to “unserved households” the signals of no more than two distant network television stations. 17 U.S.C. §§ 119(a)(2)(A)-(B). An unserved household is, with respect to a particular television network, one that is unable to receive by way of a conventional rooftop antenna a Grade B signal, as defined by the Commission, from one of that network’s affiliates. Id. § 119(d)(10)(A).
In the wake of the SHVA, the Commission adopted the ILLR computer model “to predict whether a household is likely to be able to receive a signal of the required strength” and thereby “minimiz[e] the need for on-site testing.” See Satellite Delivery of Network Signals to Unserved Households for Purposes of the Satellite Home Vieiver Act, Report & Order, 14 F.C.C. Red. 2654, ¶7 (1999) (SHVA Order ). When it was adopted, EchoStar and others criticized the ILLR model because it did not account for loss of signal strength due to variations in land cover— for example, buildings and vegetation— also known as “clutter loss.” See id. ¶ 82. The Commission acknowledged that land cover affects signal intensity, but declined to include a clutter loss factor in the model because it was “not aware of a standard means of including such information in the ILLR that has been accepted by the technical and scientific community.” Id. ¶ 83.
In response, the Congress enacted the Satellite Home Viewer Improvement Act of 1999 (SHVIA), directing the Commission to “take all actions necessary ... to develop and prescribe by rule a point-to-point predictive model for reliably and presumptively determining the ability of individual locations to receive signals [of Grade B intensity].” 47 U.S.C. § 339(c)(3). Specifically, the Commission was to “rely on the [ILLR] model set forth [in its SHVA Order ] and ensure that such model takes into account terrain, building structures, and other land cover variations.” Id. The Congress further directed the Commission to establish “procedures for the continued refinement in the application of the model by the use of additional data as it [sic] becomes available.” Id. Finally, the SHVIA amended the Copyright Act to incorporate a statutory presumption in favor of the Commission’s ILLR model. See 17 U.S.C. § 119(a)(2)(B)(ii)(I) (in copyright dispute between broadcaster and satellite carrier, court must rely upon ILLR model, as revised by the Commission over time, to
Pursuant to the SHVIA, the Commission conducted the rulemaking here under review, proposing to adjust the ILLR model to incorporate “the effects of both vegetation and buildings.” See Establishment of an Improved Model for Predicting the Broad. Television Field Strength Received at Individual Locations, Notice of Proposed Rulemaking, 15 F.C.C. Red. 1843, ¶ 9 (2000) {ILLR Notice). In particular, the Commission proposed to subtract from each predicted signal strength a “clutter loss value” based upon one of ten different categories of land cover, using data from the Land Use and Land Cover (LULC) database published by the United States Geological Survey. Id. ¶¶ 9-11. Loss of signal strength would be calculated using the figures in “Clutter Losses and Environmental Noise Characteristics Associated with Various LULC Categories,” IEEE Transactions on Broadcasting, Vol. 44, No. 3 (Sept.1988), by Professor Thomas N. Rubinstein. Id. ¶ 11. Recognizing certain limitations inherent in Professor Rubinstein’s figures, see id. ¶¶ 11-12, the Commission solicited “comment on whether other data are available that would allow [it] to expand the application of clutter loss considerations, and whether there are other approaches that are scientifically supported and could be integrated into the ILLR model to take into account losses due to vegetation and man-made structures.” Id. ¶ 11.
Various broadcasters, satellite carriers, and engineers commented upon the proposed rule. Many criticized the methodology underlying the Rubinstein data and one engineer, Richard L. Biby, submitted an alternative set of clutter loss figures. Most significant, an empirical study jointly submitted by the National Association of Broadcasters and the Association for Maximum Service Television, Inc. compared “approximately 1,000 intensity measurements,” taken during field testing in five geographic regions, with the signal strength predicted by the existing ILLR model and the model as adjusted for clutter loss based upon the Rubinstein and the Biby data. The NAB/AMST study reported for each measurement whether the various models correctly predicted the presence of a Grade B signal, predicted service at an unserved location (over-predicted), or predicted no service at a served location (under-predicted). The study concluded the proposed models were “less accurate” than the ILLR model already in use because they produced a lower percentage of correct predictions. Finally, the Associations argued that the SHVIA required the Commission to leave its “highly accurate” model in place “[a]bsent an empirically-validated method of improving the accuracy of the ILLR.” The NAB/AMST study was the only empirical study submitted by any commenter, and no commenter, including EchoStar, criticized the study prior to the Commission’s initial decision.
Relying upon the NAB/AMST study, the Commission concluded that for both VHF and UHF channels “the ILLR model without clutter corrections prove[d] superior to [the alternatives] by making the correct prediction more often.” Establishment of an Improved Model for Predicting the Broad. Television Field Strength Received at Individual Locations, First Report and Order, 15 F.C.C. Red. 12,118, ¶ 14 (2000) {ILLR Report and Order). With respect to VHF channels, the Commission concluded that reducing predicted signal strength to take account of clutter loss would “make the ILLR model less accurate because it already produces more under-predictions than over-predictions.” Id. Accordingly, the Commission set the “clutter loss values for VHF channels to zero,” id. ¶ 15, thereby leaving the results
EchoStar petitioned for reconsideration, arguing the Commission had “abdicated its responsibility” under the SHVIA to take into account clutter loss values in its predictive model for VHF channels. EchoStar also objected that the Commission had relied upon the NAB/AMST study without having made the supporting data available for comment. (The two Associations later filed those data with the Commission, approximately one month before EchoStar’s reply was due.) Finally, EchoStar argued the Commission had unlawfully refused to permit it to conduct on-site signal strength tests at the premises of customers who claim to have been erroneously identified as served, rather than requiring those customers to pursue the waiver and testing process outlined in 47 U.S.C. §§ 339(c)(2), (4), of which more later.
The Commission denied EchoStar’s petition, explaining it had not “ignore[d]” clutter loss but rather had “made a considered determination that the most accurate ILLR predictions for VHF stations under certain groundcover conditions, including buildings, are made by setting the corresponding loss values to zero.” Establishment of an Improved Model for Predicting the Broad. Television Field Strength Received at Individual Locations, Memorandum Opinion and Order, 19 F.C.C. Red. 9964, ¶ 13 (2004) (Reconsideration Order). The Commission also noted that the data underlying the NAB/AMST study “ha[d] been publicly available since well before” it issued the ILLR Notice, that the Associations had filed the data in the record since issuance of the final rule, and that EchoS-tar had “reviewed and utilized the raw data in its arguments.” Id. ¶ 10. Finally, the Commission rejected EchoStar’s argument that the statute permits it, instead of following the waiver and testing procedures in 47 U.S.C. §§ 339(c)(2), (4), to conduct its own on-site testing for customers claiming to be unserved by over-the-air signals from network affiliates. EchoStar now seeks review of the ILLR Report and Order and the Reconsideration Order.
II. Analysis
EchoStar makes three arguments: The Commission (1) by setting to zero the clutter loss term in the ILLR model for VHF stations, violated 47 U.S.C. § 339(c)(3); (2) by relying upon data not in the record, failed to comply with the notice and comment requirements of the Administrative Procedure Act, 5 U.S.C. §§ 553(b), (c); and (3) by refusing to permit EchoStar and its customers to bypass the waiver and testing process outlined in 47 U.S.C. §§ 339(c)(2), (4), violated what it identifies as a principle in the SHVA and the SHVIA that on-site testing is the only way conclusively to determine whether a household is served or unserved. We review an agency’s interpretation of a statute it administers under the familiar two-step process delineated in Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). First, we must determine whether “the intent of Congress is clear.” Id. at 842, 104 S.Ct. 2778. If so, then “that is the end of the matter.” Id. If, on the other hand, “the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether [the] agency’s answer is based on a per
A. Land Cover
EchoStar first argues the Commission failed to “ensure that [its predictive] model takes into account terrain, building structures, and other land cover variations.” 47 U.S.C. § 339(c)(3). According to EchoS-tar, the clear import of the statute is that the Congress “intended the ILLR predictive model to give dissimilar predictions based on dissimilarities in land cover”; the Commission has discretion regarding “how to incorporate land cover variations” but not regarding “whether to give them effect.” Therefore, argues EchoStar, the Commission erred when it set the clutter loss factor to zero for VHF channels rather than incorporating in the model a term that would vary with land cover.
The Commission makes two counterarguments, of which only the second is grounded in the decisions under review. First, it argues the ILLR model already takes into account land cover variation because the model “was itself derived from empirical observations of signal intensity, and those observations would themselves have reflected some degree of clutter loss.” Because the statute directs the Commission only to ensure the model “takes into account” variations in land cover — and not specifically to incorporate a variable for land cover — the Commission argues it has complied with the statute by determining that “the bias [toward under-prediction] present in the existing ILLR model has the effect of accounting for land [cover].” Though some broadcasters made this argument before the Commission, the agency never adopted it. Because we must rely only upon the reasons given by the agency, not “counsel’s post hoc rationalizations for agency action,” we disregard this ground for upholding the orders.
Second, as it did in the rulemaking, the Commission takes the position that any changes it makes to the model must comply with the statutory “command that [it] craft a reliable predictive model.” See 47 U.S.C. § 339(c)(3) (“[T]he Commission shall take all actions necessary ... to develop ... [a] predictive model for reliably and presumptively determining the ability of individual locations to receive signals of [Grade B intensity]”). According to the Commission, EchoStar’s interpretation would “read[ ] reliability out of the statute” by requiring the Commission to adopt one of the proposed adjustments even though it would reduce the accuracy of the model. Instead, the Commission says it conducted a “thorough analysis” of the adjustments proposed and “made a considered determination that the most accurate ILLR predictions for VHF stations under certain groundcover conditions ... are made by setting the corresponding loss values to zero.” Reconsideration Order, 19 F.C.C. Red. 9964, ¶ 13.
Citing our decision in Colorado v. United States Department of the Interior, 880 F.2d 481 (1989), EchoStar nonetheless argues the “Congress, not the Commission, determined that the time was right to incorporate land cover variations into the ILLR model, and the Commission had no power to second-guess that determination.” In Colorado we held unlawful a model adopted by the Department of the Interior (DOI) to assess damages to natural resources from the release of oil or hazardous substances, as required by the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA), 42 U.S.C. § 9601 et seq. Colorado, 880 F.2d at 482. Although the statute expressly provided the DOI regulations “shall take into consideration factors including, but not limited to, replacement value, use value, and ability of the ecosystem or resource to recover,” 42 U.S.C. § 9651(c)(2)(B), the DOI had promulgated a rule based only upon lost use value “because adequate data were not available to create a standardized model based on average values for restoration costs.” Colorado, 880 F.2d at 491 (quoting 52 Fed.Reg. 9051 (1987)). We rejected this justification for the Department’s failure to regulate as instructed, reasoning that “data limitations ... cannot justify DOI’s decision to ignore the clear mandate of Congress.” Id. Here, too, argues EchoStar, the agency simply ignored an express direction of the Congress because the data (in EchoStar’s words) “seemed inadequate.”
Actually, this case is quite different from Colorado. As we explained in the companion case of Ohio v. United States Department of the Interior, 880 F.2d 432 (1989), our insistence there that the model incorporate a term for restoration cost was based not merely upon its inclusion in a list of factors the regulations were to “take into consideration,” but also upon the Congress having expressed in other provisions of the CERCLA a “clear preference for restoration as the basic measure of natural resources damages.” Id. at 444-46. We thought it “would be odd indeed for a Congress so insistent that all damages be spent on restoration to allow a ‘lesser’
This conclusion does not, as EchoStar hyperbolically contends, render the mandate of § 339(c)(3) “meaningless.” That mandate is forward-looking and continuing. If EchoStar, another interested party, or the Commission itself in the future identifies an adjustment to the model that both varies with land cover and increases the accuracy of the model, then presumably the Commission will be obligated to refine the model accordingly.
B. NAB/AMST Data
EchoStar next argues the Commission violated § 553 of the Administrative Procedure Act, 5 U.S.C. §§ 553(b), (c), by failing in a timely manner to make available for public comment the raw data underlying the NAB/AMST study; indeed, the Associations did not file those data with the Commission until after the Commission had promulgated the final rule, and EchoStar had filed its petition for reconsideration and its reply was due in about one month. Although the Commission rejected this challenge in its Reconsideration Order on the ground that the underlying data were “publicly available” in federal court and in other Commission records at the time of the rulemaking, 19 F.C.C. Red. 9964, ¶ 10, EchoStar claims its efforts to obtain the data were unavailing and, in any event, it should not have been required to discover them itself. According to EchoStar, “there has never been any meaningful way for interested parties to test the data and point out errors so as to prevent the Commission from using unreliable information as the basis for its final rule.”
The Commission responds that EchoS-tar was not entitled to the data because the Commission itself neither had nor relied upon them when it issued its final rule. Rather, the Commission based its analysis upon the description, methodology, and results of the study contained in the public comments filed by the Associations. See Air Transp. Ass’n of Am. v. FAA, 169 F.3d 1, 7 (D.C.Cir.1999) (the “critical factual material that is used to support the agency’s position on review” must be made available for review).
We need not decide whether EchoStar was entitled to these data before the Commission issued its final order for the simple reason that EchoStar, although on notice of the findings and conclusions of the NAB/AMST study, did not ask for the data before the Commission issued its final rule. The record reflects the following sequence of events in 2000:(1) On February 22 the Associations submitted their comments on the ILLR Notice, in which comments they summarized the findings of their empirical study, the key finding being that the “ILLR model is already a highly accurate predictor of whether households receive— or do not receive — a Grade B signal.” (2) On March 14 EchoStar submitted its reply comments, in which it did not even mention the NAB/AMST study, and the Associations submitted their reply comments, including a chart comparing the accuracy of the various models under consideration. (3) On May 26 the Commission issued the final rule. (4) On July 10 EchoStar petitioned for reconsideration, objecting to the
EchoStar asserts “the most critical discussion of the NAB/AMSTV study ... appeared in the NAB/AMSTV’s reply comments” filed on March 14, 2000, which included the chart comparing the accuracy of the proposed modifications and of the model then in use. Because the Associations’ March 14 filing merely provided further documentation of the study they had described in their February comments, we are doubtful about the greater significance of the later filing, but we will assume as much for the sake of the argument. In any event, as the Commission points out, EchoStar did not criticize the study in any way or protest the lack of raw data in its response to the Associations’ initial comments, in which they described and relied heavily upon the study; in fact EchoStar did not even mention the study, which suggests not that EchoStar was blindsided in March but that its current objection to the unavailability of the raw data then is an afterthought.
Moreover, EchoStar could have criticized the study, or requested more time in which to do so, during the two months between the filing of the Associations’ reply comments and the issuance of the Commission’s decision. Under the agency’s liberal ex parte rules EchoStar could have submitted a written presentation at any time during the rulemaking — even, as Commission counsel said at oral argument, at the “11th hour.” See 47 C.F.R. § 1.1200, 1.1206 (in “permit-but-disclose” proceedings, including informal rulemakings, “ex parte presentations to Commission decision-making personnel are permissible but subject to certain disclosure requirements” until “the proceeding is no longer subject to administrative reconsideration or review or to judicial review”); see also Home Box Office, Inc. v. FCC, 567 F.2d 9, 57 (D.C.Cir.1977) (“[W]e recognize that informal contacts between agencies and the public are the ‘bread and butter’ of the process of administration and are completely appropriate so long as they do not frustrate judicial review or raise serious questions of fairness.... If ex parte contacts ... occur, we think that any written document or a summary of any oral communication must be placed in the public file established for each rulemaking docket ... so that interested parties may comment thereon.”). Indeed the parties to this proceeding, including EchoStar itself, repeatedly availed themselves of those rules by filing written presentations after the expiration of the formal filing date.
Arguing that disclosure of the raw data came too late because the Commission already had issued its final rule, EchoStar cites Advocates for Highway & Auto Safety v. FHA, 28 F.3d 1288 (D.C.Cir.1994), where we noted that an “agency is not likely to be receptive to suggested changes once the agency puts its credibility on the line in the form of final rules.” Id. at 1292. But it is no doubt for precisely this reason that parties avail themselves of the opportunities, ex parte or not, to object to evidence in the record before the Commission adopts a final rule. Cf. United States v. L.A. Tucker Truck Lines, Inc., 344 U.S. 33, 37, 73 S.Ct. 67, 97 L.Ed. 54 (1952) (“courts should not topple over administrative decisions unless the administrative body ... has erred against objection made
We also reject EchoStar’s argument, based upon passing references in the ILLR Report and Order to the Commission’s “further analysis” and “veriffication]” of aspects of the NAB/AMST study, that the Commission developed its own “data and analysis” upon which EchoStar should have had a chance to comment. As the Commission explains in its brief — and the record suggests nothing to the contrary — these are references merely to the agency staffs own cogitations upon the evidence in the record. Were parties entitled to comment upon every observation an agency staff member draws from the record as it accrues, rulemaking proceedings would be interminable. The APA does not contemplate so exquisite a process. Cf. Cmty. Nutrition Inst. v. Block, 749 F.2d 50, 58 (D.C.Cir.1984) (“Rulemaking proceedings would never end if an agency’s response to comments must always be made the subject of additional comments”).
C. Waiver and Testing
If the ILLR model predicts a certain household is served and its satellite carrier accordingly denies it retransmission of a distant network signal, then the satellite subscriber may ask the network station (through the satellite carrier) for a waiver of that denial. 47 U.S.C. § 339(c)(2). If the network station denies the waiver, then the subscriber may request an independent field test of signal intensity in order to show that the subscriber’s household is in fact unserved. See id. § 339(c)(4)(A). The satellite carrier and the network station or stations may jointly select a qualified and independent tester or, if they are unable to agree, then a tester will be designated by an independent and neutral entity, id. § 339(c)(4)(B), for which role the Commission has chosen the American Radio Relay League. ILLR Report and Order, 15 F.C.C. Red. 12,118, ¶ 23. The cost of the test is assigned to the satellite carrier or the broadcaster depending upon the result of the test; the loser pays. 47 U.S.C. § 339(c)(4)(B).
EchoStar contends that a satellite subscriber claiming to live in an unserved household incorrectly identified by the predictive model as being served may truncate the potentially time-consuming waiver and testing procedure specified in 47 U.S.C. §§ 339(c)(2), (4), and instead simply present to the network station the result of an independent field test showing the house is unserved. According to EchoS-tar, because the most reliable determinant of whether a person is unserved is a signal strength test, and because the “waiver/loser-pays system was adopted to ease the burden on satellite operators, who might otherwise bear the full cost of the testing,” it follows that a subscriber should be able to skip the waiver procedure by having his house tested, provided the satellite carrier agrees to pay for the test.
In support of this view, EchoStar points out that under the SHVA, “actual measurement of signal strength at the household premises [was] the only conclusive way to determine whether the household [was] unserved.” See SHVA Order, 14 F.C.C. Red. 2654, ¶ 45 (“Individual testing is the key mechanism under the SUVA for proving that a specific household is unserved”); 17 U.S.C. § 119(d)(10)(A) (defining an “unserved household” as a household that “cannot receive, through the use of a conventional, stationary, outdoor rooftop receiving antenna, an over-the-air signal of a primary network station affiliated with that network of Grade B intensity”). Therefore, reasons EchoStar, § 339 merely provides “an additional process” by which a satellite provider may establish that a household is unserved; it is more
The Commission is not unsympathetic to the policy argument underlying EchoStar’s position; it noted in its Reconsideration Order that “EchoStar has raised a valid public interest concern with the efficiency of the process used to determine SHVIA eligibility.” 19 F.C.C. Red. 9964, ¶24. The Commission further noted, however, that the “statute delineates a specific sequence of events preceding testing: waiver request, waiver denial, the subscriber’s request for an on-site test, selection of a qualified tester by the satellite carrier and the network station, and then the on-site test.” Id. ¶23. Inferring the Congress intended that the sequence detailed in the statute be exclusive, the Commission rejected EchoStar’s claim of right to a streamlined process.
We agree with the Commission. Pre-SHVIA, in the absence of any express statutory guidance to the contrary, the Commission reasonably may have concluded that individual testing was “the key mechanism” for determining whether a household was unserved, SHVA Order, 14 F.C.C. Red. 2654, ¶45, but the statute is no longer so opaque. Instead, a prediction of the ILLR model presumptively establishes whether a household is unserved, 17 U.S.C. 119 § (a)(2)(B)(ii)(I), and §§ 339(c)(2) and (4) lay out an elaborate procedure for requesting a waiver from a denial of service based upon, and for testing the result of, that prediction. Moreover, we note that EchoStar’s claim that it may circumvent the procedures in § 339 so long as it pays for an on-site measurement by an independent tester, because its right to do so pre-dates and survives the SHVIA, does not square with § 1005(a)(2) of the SHVIA, 17 U.S.C. § 119(a)(2)(B)(ii)(II). That provision directs that for “purposes of site measurements to determine whether a person resides in an unserved household under [17 U.S.C. § 119(d)(10)(A) ], a court shall rely on [47 U.S.C. § 339(c)(4) ],” which is the very provision that prescribes what EchoStar calls the “unnecessarily inefficient and burdensome” process for testing that it seeks to avoid. EchoStar points to nothing in the statute to support its conclusion that it may bypass the procedures in § 339 by conducting its own on-site testing; nor do we see how one can square the above-referenced instruction to courts with any procedure other than that in § 339.
In sum, the Commission’s reading of the SHVIA is the only reasonable one; EchoS-tar’s contrary reading finds no support in, and is actually inconsistent with, the statute. We therefore uphold the Commission’s reading under Chevron step one and end our inquiry here. See Chevron, 467 U.S. at 842,104 S.Ct. 2778 (if “the intent of Congress is clear,” then “that is the end of the matter”).
III. Conclusion
For the foregoing reasons, EchoStar’s petitions are
Denied.
The passing reference in the background section of the Reconsideration Order to the Commission's determination that it would "improve the accuracy of the ILLR model to assign certain signal loss values, in addition to those already implicit in the model” is not enough to preserve the present argument. 19 F.C.C. Red. 9964, ¶ 4. The Commission did not then, as it does now, argue that it complied with the statutory directive to take land cover into account by recognizing the ILLR model already does so.
