Here, Nestle does not dispute that if "transfer" means what the Fund
says it does, then the arbitrator would have correctly applied the law to
the facts. Nestle argues instead that the arbitrator erred in
understanding the meaning of the term "transfer" in the statute, while
the Fund argues that he got it right. We do not, therefore, have a mixed
question of fact and law, concerning the application of undisputed law to
a set of facts, id. at 1409 (quoting Pullman-Standard v. Swint,
456 U.S. 273, 289 n. 19 (1982) ("[M]ixed questions of law and fact [are]
questions in which the historical facts are admitted or established, the
rule of law is undisputed, and the issue is whether the facts satisfy the
statutory standard, or to put it another way, whether the rule of law as
applied to the established facts is or is not violated.") We have a pure
question of law to be reviewed de novo. I bear in mind that the
arbitrator has special expertise in the area of pension law, but I am as
well able to interpret statutory language as he, and I do so using
I address two preliminary matters. First, Nestle asks that I find the
arbitrator erred in refusing to strike certain evidence as irrelevant. To
the extent that certain testimony might have been irrelevant but was not
relied upon, the arbitrator did not err in admitting it to the record.
Insofar as the arbitrator relied (or I rely) on any of this evidence,
notably that concerning whom Nestle assigned to which lanes after the
closures, this was because it was thought to be relevant, as explained in
the appropriate place.
Second, the Fund argues that I should treat certain of Nestle's earlier
statements as admissions that the work was transferred within the meaning
of the statute. In particular the Fund argues that the arbitrator treated
the notes of John Holloway, Vice President for Labor Relations, made in
May 1995, as "conclusive" on this question. But the arbitrator wrote that
the notes in question "conclusively show that [Nestle] knew that the
over-the-road work of the [affected union] drivers would go to common
carriers, independent operators, and other company drivers." Whether that
constitutes a transfer within the meaning of the law is a further
The Fund also contends that it was an admission that the work was
transferred when Nestler's in-house counsel checked the "yes" box in
answer to the question whether the work had been transferred in its
Statements of Business Affairs in 1996, Def. Ex. B at 6. The arbitrator
did not in fact treat this statement as an admission, although he did
imply, without expressly so finding, that the facts suggested that Nestle
changed its tune on whether the changes in work involved a transfer when
it learned that this would mean almost a million dollars more in
withdrawal liability than it had originally anticipated. The Fund argues
that there is law suggesting that contemporaneous statements may be more
reliable than post hoc explanations when there is a motive to
misrepresent, see Mutual Life Ins. Co. v. Hillmon, 145 U.S. 285, 295
(1892), but this case is not apposite, even though Nestle points to no
relevant factual discoveries that induced it to change its mind. Nestle
is right that this is a legal question. One can admit a legal fact, but
if the law plainly indicated that no liability existed, it would be
unjust to hold Nestle to its initial word. Besides, the Fund would not
want employers to routinely deny that they ever transfer work. Therefore
I decline to find that Nestle has admitted that there was a transfer.
The principal issue is the meaning of "transfer" in § 1385(b)(2)
(A)(i) The parties say that this is a matter of first
impression, and I
cannot find any federal case law interpreting the ERISA provision at
issue, although there are several Opinion Letters by the Pension Benefit
Guaranty Corporation ("PBGC"), the agency delegated the authority to
administer the MPPAA, to which the arbitrator and the parties refer. I
must therefore proceed first from general principles of statutory
interpretation. Here the basic rule is that plain statutory language
governs, see Olander v. Bucyrus-Erie Co., 187 F.3d 599, 604 (7th Cir.
1999),*fn2 a principle rightly insisted upon by Nestle and the Fund.
Nestle's central argument is as follows. First, a PBCG Opinion Letter
87-17 (Aug. 13, 1986), Def. Ex. 3, says that work formerly done by
pension fund covered employees but taken up by third parties is not
"transfer[red] . . . to another location within the meaning of [§
1385(b)(2)(A)(i)]," because the law covers situations where "work of
the same type is continued by the employer but for which contributions to
a plan . . . are no longer required." The PBGC understood "work of the
same type" to mean in part work done "at another of its own locations."
Therefore I am to consider only the Nestle employee drivers and not the
independent owner operators. I agree with the PBGC's construction.
Second, "transfer" is a transitive verb that takes an object: someone
transfers something. It means "to convey [something] from one person,
place, or situation to another." Webster's Ninth New Collegiate
Dictionary 1253 (1990). Here, Nestle says, there was no identifiable
object. This is because the work "in the jurisdiction of the collective
bargaining agreement of the type for which contributions were previously
required," in the language of the statute, was defined by the domicile of
the drivers doing the work. With regard to the long distance lanes, there
was no way to identify any work in those jurisdictions once the St.
Joseph and Oconomowoc terminals were closed. Moreover, Nestle says,
nothing was conveyed. After the closures, the local lanes were no longer
driven by any Nestle employees, and that work was being done by third
party outsiders, and so, under the PBGC Opinion Letter, it did not involve
a transfer. And the long distance lanes driven by Nestle employees did
not increase in the number of runs after the closures; therefore, Nestle
contends, there was no transfer here either.
This is a nice piece of statutory interpretation. However, the
arbitrator correctly located the main problem with Nestle's reasoning.
This is the assumption that a transfer must involve a previously
identifiable thing transferred. The statute says no such thing. The
arbitrator noted that the work was assigned in the same way before and
after the closures, that is, on the basis of efficiency, driver
availability, location, and remaining regulated hours of the drivers.
Therefore, the arbitrator correctly found, "whether the work was
identified at all seems of little relevance. . . . Identification is
possible only by use of hindsight to see which class of drivers received
the assignment to haul freight in a particular lane."
Nestle suggests that this violates the grammar of the word "transfer,"
but English syntax does not involve the ontological commitment to a
pre-existing entity that must already somehow be if it is to be
transferred. "Work" is an abstract relation identifiable only
functionally (who has done what) and not a material substance that has an
independent existence or that can be moved from place to place like a
ball. The arbitrator was certainly correct that nothing in the law or in
the logic or grammar of the word "transfer" compels Nestle's preferred
reading in this context.
In addition, the Fund is right to point out that there is a tension
between Nestle's argument that before the closing it was not possible to
identify what work fell under the union contract, but it was possible to
say that after the closures that the remaining work was all identifiably
outside the jurisdiction of the former contract, and so had not been
transferred from covered union work. Nestle is trying to have it both
ways. The work was assigned in the same manner and was not essentially
different in character, so it was either identifiable (post hoc or ex
ante) both before and after the closures or neither before nor after.
Finally, Nestle's argument that the work was not identifiable because
it was defined by the jurisdictions of the former contracts would seem to
prove too much. If Nestle means that because the statutory language "such
work" refers to "work in the jurisdiction of the collective bargaining
agreement," as defined by the domicile of the affected drivers, but there
was no such work after the collective bargaining agreement was affected,
then the statute would be directly self-defeating: there could never be
partial withdrawal liability under this section because the termination of
the contractual obligation which is the precondition of such liability
would destroy the ability to identify the work purportedly transferred.
If, somewhat more plausibly, Nestle's point is that work assigned by
this method is not identifiable ex ante, that is correct, in the sense
that because the assignment is random, or in any case depends on
contingent factors not otherwise than merely statistically predictable,
one could not say before the assignments were actually made which work
would go to a covered union driver. However this also proves too much: if
that is a reason to deny partial withdrawal liability, then any employer
can insulate itself from such liability by the simple expedient of
adopting such a method of assigning work. I do not believe that this
reflects the intent of Congress or comports with the purpose of the
statute. Because it is also not required by the language of the statute,
I do not accept this reading.
Nestle objects that the arbitrator employed a nonstatutory presumption
in favor of a partial withdrawal because he found no evidence that "any
particular lane assignment" that went to a nonunion Nestle employee driver
after the closures would have gone to a covered union driver before.
Instead he reasoned that since the method of assigning work was the same
before and after, and since covered union drivers got some of those lanes
before, they would. have continued to get some of those lanes
afterwards. As explained above, ex ante identification of particular
lanes that covered union drivers would have been assigned is neither
possible nor necessary.
Nestle then argues that even if it is legally permissible to use :his
form of inference, the conclusion that follows is not that work was
transferred but that it merely continued elsewhere. This is supposedly
because a nonunion employee driver could perfectly well have driven any
given route before the closures as well as after. This argument not only
depends on the fallacy that work cannot be transferred unless it is
identified, it also essentially abolishes partial withdrawal liability on
transfer grounds. In fact, the arbitrator was correct to find that there
had been a transfer because, as he said, first, there were lanes driven
by covered union drivers before the closures and none afterwards, and
second, the method of assigning them was one that would have assigned
some of those jobs to covered union drivers, although it was not possible
to say which ones in advance.
Finally, Nestle argues that § 1385(b)(2)(A)(i) requires "at a
minimum" evidence that Nestle assigned nonunion Nestle employee drivers
more lanes after the closures than they were assigned before to support a
finding of transfer. However, this requirement is not found in the text
of the statute. The PBGC has indicated that there is "no quantitative
test with regard to the statutory requirement of a transfer," Op. Letter
83-20 (Sept. 2, 1983, Def. Ex. 5; on the contrary, if "any of the work is
shifted to a location where contributions are not required to the plan,
then a partial cessation of the employer's obligations will occur." Id.
(emphasis added). Nestle attempts to distinguish this opinion letter by
suggesting that it "assumes the work performed at the closed facilities
was shifted to other facilities and assumes no difficulty in identifying
the work previously performed at the closed location." The first claim is
wrong: the PBGC is setting criteria for what counts as a transfer. The
second claim is irrelevant. However the work is identified, the shifting
of any quantity of it is enough to trigger partial withdrawal liability.
Here the Fund contends that historical patterns show that over seven
quarters from 1994-95, covered union drivers averaged 8.75% of the
Oconomowoc-DeKalb lane, and the arbitrator found similar results on the
other lanes. It is reasonable to infer that there would have been about
as much work assigned to covered union drivers in the period afterwards
had the CBAs not been terminated. That work was taken up in part by
noncovered nonunion Nestle employees. There was a transfer.
That being the case, the arbitrator was correct to find partial
withdrawal liability. Because Nestle does not contest the amount, the
award is AFFIRMED.