Davis v. Signal International Texas GP, L.L.C
728 F.3d 482
5th Cir.
2013
Check Treatment
Docket

Philip A. DAVIS; Byron Day, Plaintiffs-Appellees v. SIGNAL INTERNATIONAL TEXAS GP, L.L.C.; Signal International, L.L.C.; Signal International Texas, L.P., Defendants-Appellants.

No. 12-41262

United States Court of Appeals, Fifth Circuit

Aug. 28, 2013.

728 F.3d 482

the absence of additional evidence, that the sentence affects a defendant‘s substantial rights.”75

There is “additional evidence” that Pratt‘s substantial rights were affected. The transcript of the sentencing hearing reflects that the district court chose a sentence of 87 months of imprisonment because it was within what the district thought was the correctly calculated advisory guidelines range. The district court explained to the defendant that he felt that it was appropriate to apply the guidelines and not a variance or a non-Guidelines sentence. The district court also stated on the record that it was choosing a sentence within the middle of the Guidelines range as the appropriate sentence, indicating that the Guidelines range calculated by the district court was a primary factor in the selection of the 87-months’ sentence as within a range of 78-97 months. By contrast, the 87-months’ sentence is the highest sentence under a range of 70 to 87 months’ of imprisonment, which is the highest range that the Government says applies. A middle-of-the-Guidelines sentence under a 70 to 87 months’ range would be approximately 78 to 79 months, which is 8 to 9 months shorter than the prison term selected by the district court. Moreover, we cannot say that a range of 70 to 87 months is the correctly calculated range based on the record before us. There may be issues raised when mail fraud, rather than money laundering, is used as the basis for calculating the Guidelines range that might result in a lower range than 70 to 87 months.

As discussed above, section 2E1.1 instructs the sentencing court to select a base offense level from the greater of 19 or the highest offense level of the underlying racketeering offenses. The PSR started with this base offense level and then applied a 16-level enhancement under section 2B1.1(a) based on a calculated loss of over $1 million, a 2-level enhancement under section 2B1.1(b)(8)(A) because Pratt “misrepresented that she was acting on behalf of a charitable, religious, or political organization or government agency,” and a 2-level enhancement because Pratt “abused a position of public or private trust” pursuant to § 3B1.3. The resulting a total offense level of 27 resulted in a sentencing range of 70 to 87 months, and the PSR recommended a sentence of 70 months. We are not in a position to retrace all of the steps required to determine if the PSR correctly calculated the applicable sentencing range.

* * *

Pratt‘s conviction is AFFIRMED. The sentence is VACATED and REMANDED for further proceedings.

Bryan Adam Terrell, Weller, Green, Toups & Terrell, Tommy Lee Yeates, Moore Landrey, L.L.P., Beaumont, TX, for Plaintiff-Appellee.

David Sinnott Bland, LeBlanc Bland, P.L.L.C., New Orleans, LA, Robert Prentice Vining, Charles Lee Winkelman, Attorney, LeBlanc Bland, P.L.L.C., Houston, TX, for Defendant-Appellant.1

Before REAVLEY, ELROD, and GRAVES, Circuit Judges.

REAVLEY, Circuit Judge.

This appeal involves the Worker Adjustment and Retraining Notification Act (“WARN Act“), 29 U.S.C. §§ 2101 et seq. The WARN Act requires that certain employers provide written notice within 60 days in advance of any “mass layoff” at “a single site of employment.” In 2009, Defendant-Appellant Signal International, a Gulf Coast marine services and shipbuilding company, fired a number of its workers without providing advance written notice. Plaintiffs-Appellees allege that Signal‘s reduction in employment constituted a mass layoff under the WARN Act, and thus that Signal violated the Act by failing to provide proper notice. The district court held that there was a mass layoff and entered judgment against Signal. We AFFIRM.

Defendant-Appellant Signal International1 is “a leading Gulf of Mexico provider of marine and fabrication services, including new construction, heavy fabrication, [and] offshore drilling rig and ship overhaul, repair, upgrade, and conversion.”2 From July to September 2009, Signal permanently laid off 159 of its full-time workers at its two facilities in Orange, Texas,3 including Plaintiffs-Appellees Phillip A. Davis and Byron Day.

The central issue before the district court was whether there had been a mass layoff under the WARN Act. The district court first concluded that Signal‘s two facilities in Orange, Texas, constituted a single site of employment because the facilities fell into a regulatory exception for “truly unusual organizational situations,” and thus that workforce levels were to be measured across both facilities. The district court further concluded that the proper “snapshot” for measuring Signal‘s workforce levels was July 24, 2009, the date of the first layoff alleged by Plaintiffs. Based on these two conclusions, the court ruled that there had been a mass layoff during the 90-day period following July 24, 2009. The court entered a final judgment against Signal. Signal timely appealed.

1. “Single site of employment” and the unusual organization exception

On appeal, Signal asserts two errors by the district court. Signal first argues that the district court erred in concluding that the company‘s two facilities in Orange constituted a single site of employment. Whether Signal‘s two facilities constituted a single site of employment under the WARN Act is a mixed question of fact and law. . We review the district court‘s findings of underlying fact for clear error. We review the legal question of whether there was a single site of employment based on the underlying historical facts de novo.

The WARN Act does not define what constitutes a single site of employment, but the Department of Labor‘s regulations provide guidance. See . The general rule is that “separate facilities are separate sites.” 54 Fed. Reg. 16042, 16050 (Apr. 21, 1989); see . Labor Department regulations provide some exceptions to that rule, including 20 C.F.R. § 639.3(i)(8), which states, “The term ‘single site of employment’ may also apply to truly unusual organizational situations where the [preceding paragraphs] do not reasonably apply.” The district court ruled that although Signal‘s facilities were non-contiguous, they nevertheless fell under the unusual organization exception and thus constituted a single site of employment.

There is a dearth of case law on what constitutes a “truly unusual organizational situation” within the meaning of § 639.3(i)(8). The only decision in this circuit interpreting the exception is , on which the district court relied. There we held that two facilities constituted a single site of employment where the company‘s operations were once housed together in one corporate office but were later separated due to overcrowding. However, in we relied specifically on the version of § 639.3(i)(8) from the Labor Department‘s proposed regulations, rather than the final, adopted regulations that govern the present appeal.4 Whereas the proposed regulations used the phrase “unusual organizational situations,” the final regulations use the phrase “truly unusual organizational situations.” Compare 53 Fed.Reg. 49076, 49083 (1988) (proposed regulations), with 54 Fed.Reg. 16042, 16066 (1989) (final rule), and 20 C.F.R. § 639.3(i)(8) (1995) (current version) (emphasis added). We do not need to determine the significance of that distinction, because we now rely on the current language of § 639.3(i)(8).5

The relevant facts regarding Signal‘s facilities are as follows. Since October 2008, Signal has had two facilities located in Orange, Texas, one primarily dedicated to fabrication and the other primarily to administration. Signal‘s fabrication facility covers 77 acres and is located at 91 West Front Street, where it has been located since Signal‘s incorporation. It is sometimes referred to as “the Orange yard,” or simply “the yard.”

From 2003 to 2008, Signal‘s administrative office was housed in Port Arthur, Texas, 35 miles from the Orange yard. Signal acquired the Port Arthur office pursuant to a lease that it assumed from its predecessor, which lease was set to expire in May 2008. Signal permitted the lease to expire and moved administration to Orange in order to, among other things, consolidate support and have personnel closer to the shipyard operation. Signal began housing its administrative personnel in mobile office units located at the periphery of the Orange yard. It is disputed whether the mobile units were meant to be temporary or permanent, although we agree with the district court‘s finding that Signal intended to keep its administrative employees at that location for “longer than a few months.”

About three months later, Hurricane Ike struck the Gulf Coast, causing severe damage to the Orange yard. In October 2008, from Ike and because a rental space had become available a mile away, Signal moved the bulk of its administrative employees there to what is now the Administration Building. The Administration Building is a two-story office complex located at 905 Pier Road. It is sometimes referred to as “the Administration Annex,” or simply “the annex.”

The record demonstrates that after the October 2008 move to the annex, a large number of administrative employees remained housed at the Orange yard. Rodney Meisetschlaeger, General Manager of Texas Operations, testified that “approximately two dozen” administrative employees continued to be housed at the yard after Ike. The record further demonstrates that at the time of the alleged layoffs, employees housed in the annex regularly carried out business in the yard, and vice versa. In particular, a mandatory meeting was held at the annex each Monday and was attended by many employees who were housed in the yard. Some employees who were housed in the annex would regularly—even daily—visit the yard to perform management and/or production duties. Some employees maintained offices at both the annex and the yard. It is undisputed that Meisetschlaeger, as General Manager, managed both facilities’ day-to-day operations. Signal employees who were shared by and carried out daily duties for both facilities included security, health and safety, quality control, custodial services, payroll, IT, and maintenance personnel.

Based on these facts, we conclude that Signal‘s two facilities in Orange constituted a single site of employment under § 639.3(i)(8), in particular because the situation presented here is one of the “truly unusual organizational situations” contemplated by that Labor Department regulation.6 To begin with, there was regular sharing of staff between the annex and the yard. The Labor Department has explicated that in considering exceptions to the general rule that separate sites are separate facilities, any sharing of staff between separate facilities must be regular rather than occasional. See 20 C.F.R. § 639.3(i)(3) (referring to an employer who “regularly shifts or rotates the same employees from one building to another“); (quoting § 639.3(i)(3) and stating that “occasional intermingling of various employees is insufficient to place an employer within the act‘s coverage“); see also (“[T]he essence of WARN [is] the day-to-day management and personnel.“). Here, the sharing of staff between the yard and annex was not merely occasional but in fact regular, with certain employees maintaining offices at both buildings, regular visits by personnel from one facility to the other, and use of the same security, payroll, and other staff.7

We are unconvinced by Signal‘s argument that the annex and yard had different operational purposes. Of course it is true that different units within the same operation will have different purposes if one dissects those purposes finely enough. However, what matters in determining whether separate facilities constitute a single site of employment is not the immediate purpose of this or that facility, but rather what ultimate operational purpose is served by the facilities. See § 639.3(i)(4) (“Non-contiguous sites in the same geographic area which do not share the same staff or operational purpose should not be considered a single site.” (emphasis added)); (quoting 9A Ind. Empl. Rights Man. (BNA) 595:954 (1988), which stated that “geographically separate buildings (i.e. several blocks or miles apart) would not appear to constitute a single site unless they were part of a single operation” (emphasis added)).

This is not a situation where the plaintiffs argue that Signal‘s facilities are a single site of employment simply because the facilities create the same product. See (rejecting argument that “four mines share the same ‘operational purpose’ and thus must be considered a single site,” where the four mines were independently operated with separate management and “[t]he only shared ‘operational purpose’ is that all the mines produce coal,” which “alone is insufficient to classify several non-contiguous sites as one“). Nor is this a situation where the facilities are simply local, independent sites of a large, commonly owned corporation. See (“Sites need not be contiguous in order to be considered a ‘single site‘, but in order for non-contiguous sites to be deemed a ‘single site‘, there must be some connection between the sites beyond that of common ownership.“); (holding that the St. Louis County, MO and St. Charles County, MO locations for McDonnell Douglas did not constitute a “truly unusual organizational situation” under § 639.3(i)(8) because “[a]ny connection between the two sites is nothing more than that present in most large corporate organizations“); see also 54 Fed.Reg. at 16050 (in interpreting what is now § 639.3(i)(8), stating that “individual plants should be treated individually,” especially where they span a several hundred square mile area). Here, the annex and yard constitute a single operation. Signal‘s administrative activities make sense only in relation to the fabrication activities that they support; conversely, fabrication requires the support of administration at all times. It is not as if the administration and fabrication facilities are each standalone, independent operations. Rather, they constitute one joint, integrated operation to serve the same operational purpose: the production and supply of platforms, rigs, and vessels.

Finally, geographic proximity is important. . “Contiguous facilities or those in close geographic proximity are generally single sites of employment and geographically separate facilities are generally separate sites. ... The statute and regulations plainly focus on whether the resulting job loss will be concentrated in one geographic area.” (emphasis added); see 54 Fed. Reg. at 16049 (“Even where several distinct operations are performed at a geographically connected site, that building or complex will be counted as a single site of employment.“). Here, the annex and the yard are located only one mile from each other.8 See, e.g., (holding that there was an “unusual organizational situation” where “the two sites were several miles apart“). Although we agree that the annex and the yard are non-contiguous, the two facilities are so close to one another as to be essentially one site where the joint endeavor exists.

The Labor Department included § 639.3(i)(8) as a “catchall clause” in order “to maintain some flexibility in the definition of ‘single site of employment‘, [sic] to provide for truly unusual organizational situations which [the Department] could not anticipate.” 54 Fed.Reg. at 16050. We believe that Signal‘s facilities in Orange are one such “truly unusual organizational situation.” Accordingly, the facilities constitute a single site of employment for the purposes of the WARN Act.

2. Proper “snapshot” date to determine employment levels

Next, Signal asserts that the district court erred in concluding that there was a mass layoff under the WARN Act, because the court chose July 24, 2009, as the “snapshot” date for measuring employment levels to determine whether Signal‘s reduction in force was sufficiently large to trigger the WARN Act‘s requirements—in this case, whether Signal permanently laid off at least 33% of its full-time employees. We review the district court‘s interpretation of the WARN Act and related Labor Department regulations de novo. . We review the district court‘s findings of underlying fact for clear error. See, e.g., .

According to Signal, Labor Department regulations dictate that the proper snapshot date was May 25, 2009, and thus the district court erred in using employment numbers for a different date—July 24, 2009—to conclude that there was a mass layoff. Signal‘s argument hinges on the Labor Department regulation contained in 20 C.F.R. § 639.5(a)(2), which provides:

The point in time at which the number of employees is to be measured for the purpose of determining coverage is the date the first notice is required to be given. If this “snapshot” of the number of employees employed on that date is clearly unrepresentative of the ordinary or average employment level, then a more representative number can be used to determine coverage.... A more representative number may be an average number of employees over a recent period of time or the number of employees on an alternative date which is more representative of normal employment levels. Alternative methods cannot be used to evade the purpose of WARN, and should only be used in unusual circumstances.

(Emphasis added.) Here, “the date the first notice is required to be given” was May 25, 2009, which was 60 days prior to the first alleged layoff on July 24, 2009. Therefore, Signal argues, May 25, 2009, was the proper snapshot date and the district court erred in using a different date.

The district court chose July 24, 2009, as its snapshot date based on two grounds. First, the court looked to a “very specific example” contained in the preamble to the WARN Act that is essentially identical to the case at bar—namely, where rolling layoffs are made by a hypothetical employer over a 90-day period. The Labor Department stated that in such a situation,

Assuming that no notice was given, the company is liable to all ... employees [laid off during the 90-day period] because the mass layoff threshold has been reached through separate actions which did not occur for separate and distinct causes within a 90-day period. All employees terminated within the 90-day period have suffered a mass layoff and all are entitled to 60 days’ notice before the date of their termination. For this purpose, the date on which the company size is measured is Day 1 [i.e., the day immediately preceding the first layoff].

54 Fed.Reg. at 16053 (emphasis added). The district court viewed the preamble comment as an interpretation of § 639.5(a)(2), stating that “the ... concern of the court is which controls or how should I view this. And generally when one looks at a statute, one just looks at the plain language of the statute and doesn‘t worry about other materials such as legislative history. But it turns out that with CFRs, the rule is just a little bit different.” Then, applying judicial deference to the agency‘s interpretations of its own regulations, the district court concluded that July 24, 2009, was an appropriate snapshot date, and that the employment numbers available for that date were more representative.

Second, the district court noted that § 639.5(a)(2) itself permits divergence from “the date the first notice is required to be given.” As noted above, the provision expressly provides that where “the date the first notice is required to be given ... is clearly unrepresentative of the ordinary or average employment level, then a more representative number can be used to determine coverage.” § 639.5(a)(2). The provision further provides that such “[a]lternative methods ... should only be used in unusual circumstances.” Id.. The district court reasoned that there was no evidence available to provide a snapshot of Signal‘s workforce levels on May 25, 2009. Previously, the court had excluded evidence related to Signal‘s employment levels on May 25, 2009, because Signal had presented that evidence more than two years after the start of discovery. Thus, the employment numbers available for July 24, 2009, were more representative. The court also noted that even with Signal‘s purported employment numbers for May 25, 2009, the company would be able to show only that there was a 32.5% layoff—a mere 0.5% shy of the minimum percentage required to constitute a mass layoff under the WARN Act. As the court stated, “obviously you‘re reaching and pulling and stretching for every possible number here. We‘re not involved in a case where in one case it was going to be barely 33 percent to the plaintiffs and in the other case it was just a mere 2 or 3 percent.”

We agree with the district court, especially taking both of its rationales together. As a starting point for analysis, we emphasize that Signal is not challenging the district court‘s decision to exclude Signal‘s May 25, 2009-related evidence.9 Thus, we must take as given that there was no such evidence available. See (appellant‘s failure to brief an issue on appeal constitutes waiver). This point, combined with the plain language of § 639.5(a)(2) permitting the use of an alternative date with “a more representative number,” effectively decides the issue.

Signal makes three arguments regarding the proper snapshot date: (1) that the WARN Act preamble comment was not intended to be an interpretation of § 639.5(a)(2); (2) that § 639.5(a)(2) is unambiguous (and, implicitly, that the provision unambiguously conflicts with the WARN Act preamble comment); and (3) that even if § 639.5(a)(2) is ambiguous, the preamble comment does not warrant any level of deference.10 These arguments are inapposite because they disregard the fact that § 639.5(a)(2) expressly permits divergence from “the date the first notice is required to be given,” namely where there are “unusual circumstances” such that that date is “clearly unrepresentative” of ordinary employment levels. Signal suggests by its arguments that there is some irreconcilable conflict between the language of § 639.5(a)(2) and the preamble comment. However, as demonstrated by a full reading of § 639.5(a)(2), that conflict is illusory.11

The question is not whether the district court‘s use of July 24, 2009, as a snapshot date “conflicts” with § 639.5(a)(2), but rather whether the district court was confronted with an “unusual circumstance[]” that permitted it to use “a more reasonable number” to determine coverage under the WARN Act.12 Here, it was not merely that evidence for May 25, 2009, failed to reflect ordinary employment levels at Signal; there was no such evidence for that date. The absence of employment numbers for May 25, 2009, made the “snapshot” on that date “clearly unrepresentative of the ordinary or average employment level” at Signal. § 639.5(a)(2).

Because the district court correctly concluded that the “snapshot” of May 25, 2009, was not representative of ordinary employment levels at Signal, the court was then permitted to use a different snapshot date with more representative employment numbers. Here, the district court had plenty of reasons to determine that July 24, 2009, was an appropriate snapshot date. First, the court had a specific example from the WARN Act preamble in which the Labor Department embraced the use of “Day 1” (the date immediately preceding the first layoff) as a snapshot date. Second, the July 24, 2009, date was relied on by the court and the parties for over two years; it was not until the eleventh hour that Signal began to argue for a different snapshot date and to disclose this purportedly “new” evidence that it should have disclosed at least two years earlier. Third, Signal‘s May 25, 2009-related evidence would have yielded only a 32.5% employment loss at minimum—likely higher than 33% because Signal apparently misclassified some of its laid-off employees—which thus amounted to a mere quibble over whether or not Signal fell under the requirements of the WARN Act. Taking all of this together, we conclude that July 24, 2009, was the best snapshot date. Because the parties have stipulated that there was a mass layoff during the 90-day period following July 24, 2009, the district court did not err in concluding that there was a mass layoff under the WARN Act.

AFFIRMED.

Notes

1
1. We will refer to the three Defendants-Appellants in the collective singular.
2
2. About, SIGNAL INTERNATIONAL, http://www.signalint.com/about (2012). The record also states that “Signal International is engaged in the building, refurbishing, and repair of Derricks, Platforms, Rigs, and Vessels on a project basis.”
3
3. Signal also has locations in Mobile, Alabama, and Pascagoula, Mississippi. Locations, SIGNAL INTERNATIONAL, http://www.signalint.com/contact/locations (2012).
4
4. The alleged WARN Act violations at issue in took place in 1989 when the proposed regulations were still in effect because the final regulations had not yet been adopted. See (“Because administrative rules should not be construed as having retroactive effect unless their language requires that result, ... we hold that the district court erred in applying the final regulations.“). Because the alleged WARN Act violations in the case at bar took place in 2009, the final regulations govern here.
5
5. See (unpublished) (rejecting the reasoning in on the ground that it is based on outdated regulatory language, but remarking, “We cannot say how the Fifth Circuit would have ruled had it applied the more restrictive language, but it is for us to ask whether it is ‘truly unusual’ for space constraints to cause employees originally working in a unitary facility to be redeployed under separate roofs.“).
6
6. Moreover, we conclude that the paragraphs preceding § 639.3(i)(8) “do not reasonably apply” because of the unique dynamic that existed between Signal‘s two facilities.
7
7. We do not rely on § 639.3(i)(3) or , where the court analyzed only § 639.3(i)(3) and (4), for our conclusion that the two facilities constituted a single site of employment. We conclude, however, that the sharing of staff—a factor identified in § 639.3(i)(3) and expounded upon in —is integral to our determination that the situation is “truly unusual.”
8
8. That is, as measured by the distance between the facilities’ postal addresses.
9
9. Signal explicitly states in its reply brief that it “is not appealing the district court‘s exclusion of the May 25, 2009-employment-level evidence[.]”
10
10. Signal also argues that Plaintiffs failed to present evidence at trial “establishing a base denominator number of full-time employees working at Signal‘s facilities on May 25, 2009,” and thus that they did not satisfy their burden of proof. This argument is only relevant if we agree—which we do not—that May 25, 2009, is the only appropriate snapshot date with which Plaintiffs could have made their showing.
11
11. Indeed, following the aforementioned preamble comment, the Labor Department went on to recognize that the baseline snapshot date was one among other alternative dates: “The regulation also provides for a ‘snapshot’ test for determining the number of employees in an employer‘s workforce or at a single site of employment for purposes of determining coverage. The ‘snapshot’ test is simply to look at the employer‘s employment levels on the date notice is due to be given.” 54 Fed.Reg. at 16054 (emphasis added).
12
12. There is virtually no case law interpreting what constitutes “unusual circumstances” within the meaning of § 639.5(a)(2), and the cases that exist provide little guidance. See (stating that § 639.5(a)(2) includes an “alternate method of counting to be used in ‘unusual circumstances,’ ” but providing no further analysis on that phrase), aff‘d sub nom. ; (discounting defendant corporation‘s argument that it was a cyclical business and thus entitled to use employment numbers other than those on “the date the first notice is required to be given,” because the defendant‘s supporting evidence was unreliable); see also ; . We therefore interpret the language by its plain meaning.
75
75. (emphasis omitted) (quoting ).
Read the detailed case summary