In the

    United States Court of Appeals
               For the Seventh Circuit
                   ____________________
No. 22-1122
NATIONAL LABOR RELATIONS BOARD,
                                                          Petitioner,

                               and

THEATRICAL STAGE EMPLOYEES
UNION, LOCAL NO. 2, I.A.T.S.E.
                                            Intervening Petitioner,

                                v.

JAM PRODUCTIONS, LTD., EVENT
PRODUCTIONS, INC., STANDING ROOM ONLY,
INC., and VICTORIA OPERATING CO.,
                                                      Respondents.
                   ____________________

               Application for Enforcement of an
           Order of the National Labor Relations Board.
                 No. NLRB-1, No. 13-CA-284761
                   ____________________

    ARGUED NOVEMBER 1, 2022 — DECIDED APRIL 27, 2023
               ____________________
2                                                         No. 22-1122

    Before ROVNER, BRENNAN, and SCUDDER, Circuit Judges.
    BRENNAN, Circuit Judge. Over six years ago, employees of
Jam Productions, Ltd., voted to certify the Theatrical Stage
Employees Union, Local No. 2, as their bargaining representa-
tive. Jam filed an objection to the election results, which the
National Labor Relations Board overruled. In a prior opinion,
this court granted Jam’s petition for review and instructed the
Board to hold an evidentiary hearing on the objection. Jam
Prods., Ltd. v. NLRB, 

 (7th Cir. 2018). The Board
did so, and then overruled Jam’s objection again. Now back
before this court, the Board once more seeks enforcement of
its order compelling Jam to bargain with Local 2. We discern
no reversible error in the Board’s decision to overrule Jam’s
objection and certify the election, so we grant its application
for enforcement.
                                   I
                                  A
    Jam Productions, Ltd., 1 produces and hosts live events at
venues in and around Chicago. This case centers on stage-
hands at the Riviera Theatre, one of Jam’s locations. Jam’s la-
bor needs at the Riviera vary according to its show schedule,
so it utilizes an “on-call” list to obtain stagehands. Leading up
to a performance date, Jam contacts stagehands on its call list
and secures the necessary personnel for loading and unload-
ing gear, setting up stage equipment, and operating electron-
ics, among other tasks. During the relevant period, the Riviera
Theatre call list contained approximately 55 non-union

    1 Jam Productions, Ltd., Event Productions, Inc., Standing Room

Only, Inc., and Victoria Operating Co., operate as a single employer. We
refer to them collectively as “Jam Productions” or “Jam.”
No. 22-1122                                                                3

stagehands, and Chris Shaw operated as its crew chief. We re-
fer to that group as the “Shaw Crew.”
    The Theatrical Stage Employees Union, Local No. 2, is af-
filiated with the International Alliance of Theatrical Stage
Employees and represents stagehands, including in the Chi-
cagoland area. Local 2 also operates a non-exclusive 2 hiring
hall that connects stagehands with event-related job opportu-
nities. Jobs from the hiring hall routinely pay higher wages
than other non-union work, making such referrals valued op-
portunities in the industry. Because the operation of Local 2’s
hiring hall is integral to this case, we describe it in detail.
    As indicated, the Local 2 hiring hall helps staff stagehands
to event venues in and around Chicago. The hiring hall is non-
exclusive, and it had over one thousand registered partici-
pants at the time. Those participants fall into three general
categories: Local 2 union members, non-union participants,
and union members of other locals. Interested non-union
stagehands can register with the hiring hall in a variety of
ways. Some join through formal apprenticeship programs,
while others enroll as part of organization drives. Plus, noth-
ing prevents an individual from walking into a Local 2 office
and requesting registration. To manage its participants and
allocate work, Local 2 utilizes software known as “CallStew-
ard,” which provides two main functions. First, CallSteward
inventories hiring-hall participants. Once an individual is

    2 The parties and the Board agree that the Local 2 hiring hall is non-

exclusive. As the Board correctly explains, this means Local 2’s venue
contracts do not give it “exclusive control over who will be hired for stage-
hand work.” See Alyeska Pipeline Serv. Co., 

, 126–27 (1982);
NLRB v. Teamsters “General” Local Union No. 200, 

, 784–85 (7th
Cir. 2013). Instead, the venues can hire elsewhere whenever they see fit.
4                                                  No. 22-1122

approved to start receiving referrals, a Local 2 staff member
will add that person’s name into CallSteward. A typical par-
ticipant profile includes contact information, work experi-
ence, and special skills, if any.
     Second, CallSteward allows Local 2 management to make
and track referrals to employers. When a venue requires
stagehands, it will contact Local 2 and provide details about
its event and labor needs. A Local 2 manager will then add the
event to CallSteward and start referring participants to the
job. As Local 2 staff selects stagehands, the CallSteward sys-
tem sends the stagehands a message indicating that they have
been selected for a referral and providing information about
the job. At that point, the stagehand can either accept or de-
cline the referral. Accepted referrals populate in the CallStew-
ard system, which allows Local 2 management to ensure that
venues have sufficient stagehands for their events.
    The CallSteward system has several notable features. For
instance, once a participant is registered into the system, his
or her information cannot be permanently deleted. At most, a
participant may be labeled “inactive” or “unavailable” if he
has not responded to referrals in a long time. This inevitably
means that some registered hiring-hall participants have
moved away, died, or found different work. CallSteward’s
participant cataloguing system thus makes it difficult to as-
certain precisely how many stagehands are actively seeking
work at any given time. Additionally, the system does not
show when a profile is created, but it does show when a par-
ticipant is referred. And though CallSteward helps facilitate
the process, the selection of which jobs go to which stage-
hands is not automated—a Local 2 staff member must manu-
ally assign referrals. During the relevant time, that
No. 22-1122                                                    5

responsibility fell on two Local 2 employees: Thomas
Herrmann and occasionally Craig Carlson. Those men
wielded significant power because the number of hiring-hall
participants looking for work perpetually outnumbered
available jobs. Indeed, the Board recognized “that the number
of participants (more than 1,000) exceeded the number of jobs
to be filled (hundreds), even on the busiest days.” So,
Herrmann and Carlson controlled which participants re-
ceived the limited number of referrals each day.
    When making allocations, Herrmann and Carlson always
retained personal discretion, taking into consideration a num-
ber of factors. One important factor was employer preference.
Contracting venues would frequently request specific hiring-
hall personnel for their events and, in those cases, Local 2 staff
would try to honor their requests. Relatedly, employers some-
times asked for stagehands with unique skills, such as light-
ing or rigging ability. When possible, the hiring hall would
provide stagehands who could render the needed services.
Herrmann and Carlson would also try to preference Local 2
members over non-members or affiliate local members. In ad-
dition they also considered pragmatic aspects, like amount of
work required, stagehand availability, general experience lev-
els, and the preference for an experienced stagehand to be on
each job site as a leader. General fairness also played a role.
Still, Local 2 staff followed no strict protocol (such as worker
seniority) when making referrals. Instead, they retained com-
plete discretion over which stagehands received which jobs,
and no stagehand was guaranteed to receive any particular
hiring-hall referral.
  But at least one factor bearing on referrals fell outside
Herrmann and Carlson’s control: the weather. Given the
6                                                           No. 22-1122

nature of stagehand work in Chicago, there are routinely
more large-scale events, and thus more stagehand job oppor-
tunities, available during the warmer spring, summer, and
fall months than in the winter. And not every year is the same
as the last. The record shows that the months surrounding the
representation election in the spring of 2016 were extraordi-
narily busy for the hiring hall.
    With this context, we briefly survey the lengthy dispute
between Jam and Local 2. This case traces back to September
2015 when Jam collectively fired Chris Shaw and the entire
Shaw Crew just as they were on the brink of unionizing, re-
placing them with a new list of on-call workers. The day fol-
lowing the discharge, Local 2 filed a representation petition 3
and, shortly thereafter, submitted an unfair labor practice
charge against Jam as well. The Board’s Regional Director or-
dered the representation petition held in abeyance pending
resolution of the unfair labor practices charge. In the mean-
time, Jam and Local 2 agreed to provisional rules for a future
representation election. Specifically, the parties deemed 21
members of the Shaw Crew eligible to (eventually) vote, cre-
ating a subset we refer to as the “Shaw voters.” The parties
continued to litigate the unfair labor practices complaint 4
through the winter of 2016. Many of the fired Shaw Crew
members were able to find work during this time through re-
ferrals from the Local 2 hiring hall.


    3 The petition also sought to represent stagehands at the Vic Theatre

and Park West Theatre, both Jam venues.
    4 After investigating Jam’s charge, the Regional Director issued a com-

plaint alleging the firings violated the National Labor Relations Act, 

 et seq.
No. 22-1122                                                   7

    On March 28, 2016, Jam and Local 2 settled the unfair labor
practices complaint, with the Regional Director approving the
settlement over Local 2’s objection. In short, Jam promised to
put the Shaw Crew back on its call list and agreed to provide
the fired crewmembers stipulated backpay. Yet the settlement
did not entirely restore the pre-termination status quo for the
Shaw Crew. As noted, Jam constructed a new call list shortly
after terminating the Shaw Crew stagehands. Post-settlement,
those replacement stagehands did not simply disappear. In-
stead, the Shaw Crew had to share job opportunities with the
alternative group of workers. This resulted in Shaw Crew
members receiving significantly fewer Riviera Theatre job
offers than before they were fired. Given their increased avail-
ability, many Shaw Crew members continued to obtain refer-
rals through the hiring hall into spring 2016. So, by the time
of the election, every Shaw Crew member who was eligible to
vote and able to work had accepted Local 2 hiring-hall refer-
rals.
    Here, we pause to identify several pertinent time periods
leading up to the election. The first is known as the “critical
period,” and it spans from the filing of the election petition
(September 17, 2015) to election day (May 16, 2016). In labor
law, the critical period is typically the window of time in
which the Board looks for objectionable election conduct. See
Ideal Elec. & Mfg. Co., 

, 1278 (1961); NLRB v.
Wis-Pak Foods, Inc., 

 (7th Cir. 1997). However,
Local 2’s unfair labor practice charge against Jam—and the
ensuing complaint—created an “abeyance period” from mid-
September 2015 to late March 2016, disrupting the ordinary
operation of the critical period. As a consequence, the parties
and the Board identify a third, narrower window they call the
“focal period.” The focal period covers the six weeks
8                                                             No. 22-1122

immediately preceding the election, running from April 1,
2016, to May 16, 2016. We continue this convention, referring
when appropriate to the critical period, abeyance period, and
focal period.
    The election was held on May 16, 2016. Local 2 prevailed
on election day, but Jam promptly objected to the outcome.
Per Jam, Local 2 improperly influenced the election by provid-
ing “economic benefits to employees to induce them to
support the Union.” Namely, Jam accused Local 2 of prefer-
entially providing Shaw voters with lucrative job referrals in
the weeks and months leading up to the election. 5 The Re-
gional Director overruled Jam’s objection without a hearing,
and the Board denied review. Seeking federal court involve-
ment, Jam refused to bargain. See NLRB v. City Wide Insulation
of Madison, Inc., 

, 657 n.1 (7th Cir. 2004) (explain-
ing that certification orders are not independently appealable
to federal courts, but an employer can obtain judicial review
“by refusing to bargain and then asserting its objections to the
election as a defense to the ensuing charge of an unfair labor
practice”); see also 

(f). The Board’s general
counsel filed an unfair labor practice complaint, which the
Board resolved on summary judgment against Jam. Jam then
filed a petition for review in this court, and the Board filed a
cross-application for enforcement. Jam Productions, 

.




    5 As noted, other voters participated in the representation election, but

we focus exclusively on the Shaw voters. The Shaw voters are sufficiently
numerous on their own to have a material impact on the election, and
Jam’s objection pertains only to them.
No. 22-1122                                                    9

                               B
    This court resolved Jam’s initial appeal by declining to en-
force the Board’s certification order and remanding for addi-
tional factfinding. 

 at 1045–47. Our main concern was the
Board’s decision not to hold an evidentiary hearing on Jam’s
objection. We acknowledged that “[t]he Regional Director is
obligated to hold a hearing only when the objecting party
raises ‘substantial and material factual issues’ sufficient to
support a prima facie showing of objectionable conduct.” 

at 1044 (quoting Clearwater Transp., Inc. v. NLRB, 

 (7th Cir. 1998)). But we concluded that Jam had
made the requisite showing. 

 We explained, “The financial
benefit of the higher-paying jobs immediately preceding the
election could plausibly be seen as an economic inducement
to secure votes in favor of Local No. 2.” 

. And deter-
mining whether Local 2 had improperly provided jobs to vot-
ers could only be accomplished by granting Jam subpoena
power and an evidentiary hearing: “[W]hether the jobs were
in fact offered to the Shaw crew ‘according to pre-existing
standards and practice’ is precisely the question Jam sought
to answer with its objection.” 

 Finally, we recognized, “If
the jobs were in fact, as Jam maintains, previously unavailable
to those [Shaw crew] employees, the offer of the premium-
pay jobs could certainly be seen as an unearned benefit to in-
duce union support.” 

. On that basis, we returned
the case to the Board.
    On remand, the Board reopened the case and allowed Jam
to subpoena hiring hall records from Local 2. The Board then
held a three-day hearing during which a hearing officer lis-
tened to witness testimony and considered several statistical
summaries extrapolated from Local 2’s hiring hall data.
10                                                            No. 22-1122

Because this data forms the foundation of Jam’s arguments on
appeal, we highlight the main points of the voluminous evi-
dentiary hearing record and include noteworthy charts when
helpful. For ease of reference, we also assign numbers to the
data sets.
    Data Set One. Recall that Shaw voters are the voter-eligible
subset of the Shaw Crew. The first data set traces the number
of hiring-hall referrals given to Shaw voters over time. Before
the critical period, only one Shaw voter received referrals.
From the critical period to the start of the focal period, 17 of
20 6 Shaw voters received 107 referrals—an average of 6.3 re-
ferrals per voter. Between the start of the focal period and the
election, all 20 Shaw voters received referrals. The voters re-
ceived 213 total referrals during that time—an average of
10.65 per voter. 7 Therefore, the data indicates a general trend
of more referrals as the election approached. More voters re-
ceived referrals, and the per-voter rate of referral increased.




     6 Though 21 Shaw Crew members were eligible to vote, one member

was not fit to work at the time.
     7 The parties differ slightly on how many referrals Shaw voters re-

ceived during the focal period. Jam identifies 213 referrals that, across 20
Shaw voters, results in a per person average of 10.65 during the focal pe-
riod. The Board identifies Jam voters as receiving 215 referrals during the
focal period. Yet, in its appellate brief, the Board continues to utilize the
10.65 per person rate, a number consistent with 213 referrals divided by
20 stagehands. We note this distinction for purposes of clarity, but it does
not impact our analysis. Moreover, the record appears to support both
counts depending on which chart is consulted.
No. 22-1122                                                  11

                         Data Set One
                                                   Referrals
    Period               Voters      Referrals
                                                   Per Voter
 Pre-Critical Period
 (Pre-September 17,       1/20           2             2.0
 2015)
 Start of Critical Pe-
 riod (September
 17, 2015) to Start of    17/20         107            6.3
 Focal Period
 (March 31, 2016)
 Start of Focal Pe-
 riod (April 1, 2016)
                          20/20         213          10.65
 to Election Day
 (May 16, 2016)


    Data Set Two. A second data set compares focal period re-
ferrals for 14 Shaw voters who received their first referral be-
tween September 2015 and May 2016 to 111 other, non-Local
2 stagehands who also started receiving referrals in the same
period. But while this data set captures participants with a va-
riety of first-referral dates, it includes referral numbers only
for the focal period. As such, it cannot be used to examine
change in referral patterns over time between the Shaw voters
and the other non-Local 2 participants. It does show that dur-
ing the focal period, the 14 relevant Shaw voters received 129
referrals—an average of 9.21 per voter. In the same period, the
111 non-Local 2 stagehands received 294 referrals—an aver-
age of 2.65 per person. We need not delve into detail, but at
12                                                           No. 22-1122

times the parties and the Board slice this data set into even
more pieces.


                              Data Set Two

                                                Other Non-Local 2
                    Shaw Voters with
                                                 Participants with
                    First Referral Sept.
                                                First Referral Sept.
                     2015 - May 2016
                                                  2015 - May 2016

 Number of
                              14 8                        111 9
 Stagehands

 Focal Period
                              129                         294
 Referrals

 Focal Period
 Referral Per                9.21                         2.65
 Person


    Data Set Three. A third data set compares Shaw voters to
all other registered non-union hiring-hall participants during

     8 The Board equivocates on the number of Shaw voters who received

their first referral between September 2015 and May 2016, suggesting it is
14 or 15. This distinction is not material to our analysis.
     9 We were unable to replicate the 111 “other participant” figure in our

examination of the record. By our count, there were 145 hiring-hall partic-
ipants who began receiving referrals between September 2015 and May
2016. Removing the 14 Shaw voters from that count leaves 131 other par-
ticipants, who collectively received 294 referrals. Even if the greater num-
ber is correct, our conclusions remain the same.
No. 22-1122                                                 13

the focal period. For that stretch, 827 non-union participants
received 4,386 total referrals—an average of 5.3 per person.
The twenty Shaw voters received 213 referrals during that
time—10.65 per person. That said, the record indicates that
only 314 of the 827 non-union participants received referrals
during the focal period. Thus, the per person rate for non-un-
ion participants who actually received referrals during the fo-
cal period is 13.97.


                        Data Set Three

                                  Registered     Referred
                 Shaw Voters     Non-Union      Non-Union
                                 Stagehands     Stagehands


 Number of
                      20             827             314
 Stagehands


 Focal Period
                     213             4386           4386
 Referrals


 Focal Period
 Referrals Per       10.65           5.3            13.97
 Person



    Data Set Four. A fourth data set compares referrals given
to Shaw voters against those given to non-voter Shaw Crew
members. While other data sets lack referral data outside the
focal period, this set better demonstrates change over time.
14                                                    No. 22-1122

From the start of the critical period to the start of the focal pe-
riod, Shaw voters received 5.35 referrals per person. During
the same period, non-voter Shaw Crew members received
0.95 referrals per person. Within the focal period, Shaw Crew
voters received 10.65 referrals per person, while the non-vot-
ers received 3.3. Thus, while Shaw Crew non-voters routinely
received fewer total referrals, their per person rate of referral
also increased as the election approached.


                          Data Set Four
                 Critical Period    Focal Period      Referral
     Group       to Focal Period      Referral        Rate %
                  Referral Rate        Rate           Change
 Shaw Crew
                        5.3               10.65        200.94%
 Voters
 Shaw Crew
                       0.95                3.3         347.37%
 Non-Voters


    Post-Election Data & Grouping Data. The record touches on
two additional areas of importance: referral patterns post-
election; and referral grouping patterns. Starting with post-
election data, in the first six weeks following the election,
Shaw voters saw a 28 percent decrease in referrals compared
to the focal period. But in the next six-week period, Shaw vot-
ers received 35 percent more referrals than during the preelec-
tion focal period. Said another way, referrals to Shaw voters
dipped after the election but picked up again later. As to ir-
regular grouping, Jam identified several occasions in which
Shaw voters were assigned to work the same events together
in a way that could not be plausibly attributed to completely
No. 22-1122                                                 15

random assignment. Jam also emphasizes that Justin Huff-
man, a Shaw voter and Local 2’s observer during the election,
received the most referrals of any Shaw voter in the critical
and focal periods.
    Other Evidence. Aside from statistics, the Hearing Officer
heard testimony from several witnesses, including Local 2 call
steward Thomas Herrmann and Local 2 business manager
Craig Carlson—the men responsible for assigning referrals
during the critical and focal time periods. We expand on their
testimony when relevant below, but those men denied un-
fairly preferring the Shaw voters when assigning job referrals.
After considering the data sets and other evidence, the Hear-
ing Officer recommended overruling Jam’s objection, and the
Regional Director agreed. The Board granted Jam’s request
for review.
    After considering the evidentiary record, the Board
largely concurred with the Regional Director and overruled
Jam’s objection. In reaching its decision, the Board detailed
how cases featuring hiring-hall referrals—like Jam’s—should
be analyzed to determine objectionable conduct. The Board
started with the rules generally applicable to the preelection
context, recognizing that while “[a] union cannot make, or
promise to make, a gift of tangible economic value as an in-
ducement to win support in a representation election,” “[n]ot
every grant during an election campaign requires a ‘per se
finding’ of objectionable conduct.” What matters is “whether
the donor’s conduct would reasonably have a ‘tendency to in-
fluence’ the outcome of the election.”
   To answer that question, the Board explained it has often
employed the approach from its B&D Plastics, Inc., 

, 245 (1991), decision. There, the Board examined a set of
16                                                 No. 22-1122

factors to evaluate whether the benefit granted had the ten-
dency to influence the election’s outcome. 

 Those factors in-
cluded the size of the benefit in relation to the reason for
granting it, the number of employees who received it, and
“how employees reasonably would view the purpose of the
benefit.” 

 Then, to determine whether the grant of benefits
was objectionable, the Board drew a rebuttable inference that
the grant of benefits during the critical period was coercive.
That inference can be overcome through “an explanation,
other than the pending election, for the timing of the grant or
announcement of such benefits.” 

    Here, the Board examined the approach from B&D Plastics
and reasoned that circumstances involving employment- or
hiring-hall-related benefits introduce additional complexity.
Those types of benefits “raise a different set of analytical con-
cerns,” from other types of benefits, as they “do not involve
mere ‘incidents’ or benefits of union membership but, instead,
implicate employees’ access to employment and related bene-
fits.” Employment-related benefits have also, according to the
Board, received different analytical treatment. The Board
observed that “in hiring-hall cases, [it] has not expressly con-
sidered whether a union’s promise or grant during the critical
period may raise an inference of coercive timing, or whether
any such inference could be rebutted.” Nor has it always
“weigh[ed] the other B&D Plastics factors.” Instead, it has of-
ten “focused on whether the employees in question would
have been entitled to the benefit under the union’s normal
practices.”
   Against that backdrop, the Board employed portions of
the B&D Plastics framework alongside inquiry into normal
union practices. Namely, the Board articulated a “reconciled”
No. 22-1122                                                   17

approach for cases involving employment- and hiring hall-re-
lated benefits: “Where an objecting party alleges a union
granted access to a hiring-hall benefit during the critical pe-
riod, it has the burden of proving not just that the union did
so but also that the benefit was one to which the employees
were not otherwise entitled.” If the objecting party meets that
burden, the Board will draw an inference that the benefits
granted are coercive. It is then on the responding party to
demonstrate that the benefits were conferred for reasons un-
related to the pending election.
    Given all this, the Board scrutinized the facts of Jam’s case
and overruled its objection for two main reasons. First, the
Board determined that the referrals were conferred in the
regular operation of the hiring hall and, as such, did not con-
stitute an objectionable benefit. So, Jam failed to raise an in-
ference of coercion at the outset. Second, the Board held
that—even if the hiring-hall referrals qualified as an objection-
able benefit raising an inference of coercion—Local 2
provided persuasive, non-election-related reasons for the in-
creased number of referrals. The Board concluded that any in-
creases in referrals were attributable to the collective firing of
the Shaw Crew and the busy spring 2016 season. Part of the
Board’s decision relied on “Table A,” which it created using
native hiring-hall referral data. We reprint that graphic below:
18                                                    No. 22-1122




In addition to overruling Jam’s objection, the Board also cer-
tified Local 2 as the employees’ bargaining representative for
the Riviera Theatre and the two smaller venues.
    Following the Board’s decision, Jam again refused to bar-
gain so that it could access judicial review. See NLRB v. Amer-
iCold Logistics, Inc., 

 (7th Cir. 2000) (“Refusing
to bargain is the only way for an employer to get judicial re-
view of an NLRB decision upholding an election and certify-
ing a union.”) (citation omitted). The Board’s general counsel
filed an unfair labor practices complaint alleging Jam’s refusal
to bargain violated § 8(a)(1) and (5) of the National Labor Re-
lations Act, 

(a)(1), (5), and moved for summary
judgment. The Board granted that motion on January 11,
2022, ruling that Jam’s refusal constituted an unfair labor
practice. The Board then docketed this application for en-
forcement on January 25, 2022, and Local 2 filed a supporting
brief as intervenor. Jurisdiction over the petition for enforce-
ment is proper under 

(e). We now review the
Board’s determinations.
No. 22-1122                                                   19

                               II
    The National Labor Relations Act forbids a union from
“both blatantly giving something of value to an employee in
exchange for his vote as well as offering a benefit in a way that
‘tacitly obliges the employee’ to vote for the union.” Jam Pro-
ductions, 

 (quoting Freund Baking Co. v. NLRB,

, 931 (D.C. Cir. 1999)). Therefore, “[i]n consider-
ing whether a particular incentive taints the fairness of the
election, we ask whether what is offered is ‘sufficiently valu-
able and desirable in the eyes of the person to whom they are
offered, to have the potential to influence that person’s
vote[.]’” Id. (quoting NLRB v. River City Elevator Co., 

 (7th Cir. 2002)). That said, Congress has entrusted
the National Labor Relations Board with “wide discretion to
ensure the fair and free choice of bargaining representatives.”
NLRB v. Savair Mfg., Co., 

, 276–77 (1973) (quoting
NLRB v. Wyman-Gordon Co., 

 (1969)). So,
when a party objects to a union representation election, the
Board bears initial responsibility for investigating and deter-
mining whether the complained of conduct substantially im-
paired the exercise of free choice such that a new election
must be held. River City Elevator, 

; see also Jam
Productions, 

.
    Our subsequent review of a Board’s decision to certify an
election is deferential. “We presume the validity of a Board-
supervised election and will affirm the Board’s certification of
a union if that decision is supported by substantial evidence.”
AmeriCold Logistics, 

 (citation omitted); 

(e) (“The findings of the Board with respect to
questions of fact if supported by substantial evidence on the
record considered as a whole shall be conclusive.”). The
20                                                   No. 22-1122

substantial evidence standard is not arduous. “Substantial ev-
idence is ‘such relevant evidence as a reasonable mind might
accept as adequate to support the conclusion of the Board.’”
NLRB v. Mickey’s Linen & Towel Supply, Inc., 

(7th Cir. 2006) (quoting SCA Tissue N. Am. LLC v. NLRB, 

 (7th Cir. 2004)). And the “burden is on the ob-
jecting party to prove that the election is invalid.” NLRB v.
WFMT, 

 (7th Cir. 1993) (citing NLRB v. Serv.
Am. Corp., 

 (7th Cir. 1988)).
    “We ‘do not reweigh the evidence,’ and the ‘presence of
contrary evidence does not compel us to reverse the Board’s
order.’” ADT, LLC v. NLRB, 

, 987 (7th Cir. 2022)
(quoting Contemp. Cars, Inc. v. NLRB, 

, 868–69 (7th
Cir. 2016)). “Our only task is to evaluate ‘whether there is ev-
idence in the record supporting the Board’s outcome that
would satisfy a reasonable fact finder.’” Mondelez Glob. LLC v.
NLRB, 

, 769 (7th Cir. 2021) (quoting NLRB v. KSM
Indus., Inc., 

 (7th Cir. 2012)). This means that
a party objecting to the results of a Board supervised election
faces “a formidable burden … to prove that it was not valid.”
River City Elevator, 

 (citing Service American
Corp., 

); see also NLRB v. Erie Brush & Mfg. Corp.,

 (7th Cir. 2005) (“Our review of the Board’s
decision to certify a collective bargaining agent following an
election is extremely limited.”) (citation omitted).
   Judicial scrutiny of the Board’s treatment of substantive
labor law is similarly deferential. ADT, 54 F.4th at 987. “Con-
gress intended to confer upon the Board broad authority to
develop national labor policy and so we will uphold the
Board’s legal conclusions so long as they have a reasonable
basis in the law.” Erie Brush, 

 (citation omitted);
No. 22-1122                                                   21

see also City Wide Insulation, 

 (“Our function is
to decide whether the Board’s … legal conclusions have a rea-
sonable basis in law.”) (citation omitted). Put another way, we
accept the Board’s legal conclusions “unless they are irra-
tional or inconsistent with the [Act].” Loparex LLC v. NLRB,

 (7th Cir. 2009) (alteration in original) (quot-
ing Ryder Truck Rental v. NLRB, 

 (7th Cir.
2005)). Indeed, the Supreme Court has “refused enforcement
of Board orders where they had ‘no reasonable basis in law,’
either because the proper legal standard was not applied or
because the Board applied the correct standard but failed to
give the plain language of the standard its ordinary mean-
ing.” Ford Motor Co. v. NLRB, 

 (1979) (citation
omitted). With that deferential standard in mind, we turn to
the Board’s decision.
                               A
    First, we consider the Board’s approach to reviewing hir-
ing-hall referrals granted in the critical period. Our analysis
proceeds in two steps. We survey how the Board and this
court have addressed the promise or grant of benefits within
the critical period. Then, relying on that precedent, we deter-
mine whether the Board reasonably applied the law through
its articulated legal framework.
   As previously stated, the Board decision identified an ap-
proach applicable to employment- and hiring-hall-related
benefits conferred during the critical period. The Board ex-
plained: “Where an objecting party alleges a union granted
access to a hiring hall benefit during the critical period, it has
the burden of proving not just that the union did so but also
that the benefit was one to which the employees were not oth-
erwise entitled.” As the Board wrote:
22                                                 No. 22-1122

       [O]bjecting parties can meet this burden in mul-
       tiple ways, including by showing that the eligi-
       ble voters received favorable treatment, that the
       grant of benefits deviated from the status quo or
       the union’s normal practice, or that the eligible
       voters were treated differently than others with
       access to the referral system, among others.
If the objecting party makes that showing, the Board will
draw an inference that the benefits granted were coercive. The
responding party can then rebut the inference by explaining
the reason for the timing, which must not be election-related.
    On appeal, Jam takes issue with this framework and urges
us to find it unreasonable. Jam offers two arguments. To
begin, Jam contends that the Board’s approach imposes “sua
sponte an unprecedented two-part burden on Jam.” Jam must
demonstrate the union provided employees a benefit and that
the benefit was one to which the employees were “not other-
wise entitled” to receive. But Jam’s rejection of the Board’s
rule goes further. Per Jam, any union grant of benefits during
the critical period should automatically give rise to an infer-
ence of objectionable conduct: “[T]he Board improperly de-
parted from well-established precedent … that granting a
benefit during the critical period is, itself, sufficient to create
an inference of objectionable conduct.” Under Jam’s interpre-
tation, any benefit conferred during the critical period would
give rise to an adverse inference. It would not matter whether
that benefit was one to which the employee was already enti-
tled.
    We begin with the Board’s statement that a party objecting
to employment- or hiring hall-related benefits must demon-
strate both the delivery (or promise) of such benefits and that
No. 22-1122                                                      23

the employees were “not otherwise entitled” to the benefits.
Both this court and the Board have used a variety of ap-
proaches when evaluating whether a grant of benefits during
the critical period is objectionable. A tour of some of those ap-
proaches is instructive.
    1. The Board’s analysis in B&D Plastics demonstrates one
prominent approach. 302 NLRB at 245; see also Gulf State
Canners, 

, 1327 (1979) (identifying the factors
recited in B&D Plastics). In B&D Plastics the Board used a set
of factors to determine whether a benefit would tend to un-
lawfully influence an election. 302 NLRB at 245. Specifically,
the Board analyzed “(1) the size of the benefit conferred in re-
lation to the stated purpose for granting it; (2) the number of
employees receiving it; (3) how employees reasonably would
view the purpose of the benefit; and (4) the timing of the ben-
efit.” Id. On the timing element, the Board explained it infers
that benefits granted during the critical period are coercive.
Id. That inference is rebuttable, however, if the party responds
with an explanation unrelated to the election. Id.
   To determine when a grant of a preelection benefit is po-
tentially objectionable, several Board decisions deploy the
B&D Plastics factors and draw an adverse inference about
benefits conferred in the critical period. See, e.g., Star, Inc., 

, 962–63 (2002); Va. Concrete Corp., 

,
1182, 1184 (2003); Cmty. Options NY, Inc., 

,
1535–36 (2013); BFI Waste Sys., 

, 935 n.3, 935–36
(2001).
    2. The B&D Plastics framework is often applied with var-
iations. For instance, some Board decisions borrow the B&D
Plastics factors, but do not consider any inferences from
24                                                  No. 22-1122

timing. See, e.g., Broward Cnty. Health Corp., 

,
212–13 (1995); Sequel of N.M. LLC, 

, 1124–25
(2014).
    3. Still other Board decisions take the inverse approach,
inferring that the grant of benefits during the critical period is
coercive and thus objectionable, without analyzing any other
factors. Mercy Hospital Mercy Southwest Hospital, 

 (2002), is one such case. There, an employer announced a
wage increase during the preelection critical period. Id. at 545.
Explaining that “[t]he standard for determining whether the
timing of benefit announcement during the critical period is
unlawful is essentially the same as the standard for determin-
ing whether the grant of benefit itself violates the Act,” the
Board inferred coercive conduct from the announcement. Id.
But the Board did not explicitly evaluate the announced wage
increase using the B&D Plastics factors. Instead, the Board fo-
cused its analysis on whether the employer had provided an
explanation, other than the pending election, for the timing of
the benefits announcement. Id. Finding that the employer
failed to justify the timing of the wage increase, the Board set
aside the election. Id. at 546.
    Additional Board decisions utilize the same approach. See,
e.g., SBM Mgmt. Servs., 

, 1207 (2015) (providing
no multifactor analysis but holding that an employer grant of
bonuses during the critical period raised an inference of coer-
cion, which the employer failed to rebut); United Airlines
Servs. Corp., 

, 954–55 (1988) (ordering a hearing
for an objection and explaining that “[i]n determining
whether a grant of benefits is objectionable, the Board has
drawn the inference that benefits granted during the critical
No. 22-1122                                                   25

period are coercive, but it has allowed the employer to rebut
the inference”) (citation omitted).
    4. Further Board decisions—many of which deal with
employment- or hiring hall-related benefits—set aside factors
and inferences entirely, and instead ask whether a party con-
ferred a benefit to which the employees were not otherwise
entitled to receive. For instance, in Mailing Services, Inc., 

, 565 (1989), a union announced and provided free
medical screenings to voting employees just days prior to the
representation election. 

 Reviewing that conduct, the Board
explained that unions, like employers, are “barred in the crit-
ical period prior to the election from conferring on potential
voters a financial benefit to which they would otherwise not be
entitled.” 

 (emphasis added) (citing McCarty Processors, Inc.,

, 703 (1987)). And, because the union there
“made no contention that the employees who received the
screenings were entitled to receive them independent of the
election campaign,” 

 the Board sustained the employer’s
objection and set aside the election. Id. at 566.
    The Board used a similar approach in IBEW Local 103
(Drew Electric), 

 (1993). There, a union operated
a hiring hall through which referrals were made for construc-
tion and residential contracts. Construction contract referrals
were more lucrative, but they required the job seeker to hold
an “A” card from the union. 

 By contrast, residential con-
tracts entailed lower wages but could be obtained with either
an “A” or an “R” card. 

 Predictably, the “A” cards required
“superior qualifications and work experience” compared to
“R” cards. Id. at 592. Leading up to a representation election,
the union promised all bargaining unit employees that they
would immediately receive “A” cards if the union prevailed.
26                                                 No. 22-1122

Id. Considering the valuable positions that an “A” card would
unlock, that promise implicated a significant benefit. While
the employer objected, the Board sided with the union, find-
ing it reasonable to infer that the employees already had the
“necessary skills to qualify for ‘A’ cards.” Id. at 592. So, the
union was not promising the voters anything that they were
not already qualified for and entitled to have. Id. at 593.
    The Board reached a different result in Alyeska Pipeline, 261
NLRB at 127, while employing a similar approach. There, the
Board set aside an election because the union promised voters
undeserved—and indeed unlawful—special treatment. Id.
The union in Alyeska Pipeline ran an “exclusive” hiring hall,
which legally constrained its discretion when handing out
work. Id. Nonetheless, the union promised voters that mem-
bership would provide a “definite advantage” in hiring-hall
referrals. Id. at 126. The Board found that communication ob-
jectionable because it promised, in essence, a hiring-hall perk
that the employees were not entitled to receive; the union had
“clearly promised to give members an unlawful advantage.”
Id. at 127. The “not otherwise entitled to” rubric appears in a
collection of other Board decisions. See Topside Constr., Inc.,

, 886, 898–99 (1999); Go Ahead N. Am. LLC, 

, 77–78 (2011); McCarty Processors, 286 NLRB at 703.
    Board decisions aside, this court has at times reviewed
election-related benefits by asking whether employees re-
ceived a benefit outside their employment status quo—that is,
a benefit to which they were not otherwise entitled. In NLRB
v. Chicago Tribune Co., a union provided credential cards to
certain employees during the critical preelection period. 

, 793, 796–97 (7th Cir. 1991). The Board determined
that the union’s grant of credential cards did not improperly
No. 22-1122                                                      27

influence the election, and we agreed. Id. at 796–97. In so de-
ciding, this court explained that a party challenging an elec-
tion “must show that the unlawful acts occurred and ‘that
those acts interfered with the employees’ exercise of free
choice to such an extent that they materially affected the re-
sults of the election.’” Id. at 794 (quoting Service American
Corp., 

). But the union there gave the credential
cards according to a routine schedule separate and apart from
the election. Id. at 797. Thus, “because there was no change in
the status quo with regard to employee benefits,” the creden-
tial cards did not interfere with the election. Id.
    We relied on a comparable approach in River City Elevator
but reached the opposite conclusion. 

. There,
a union made a pre-election promise that all employees
would receive mechanics cards, even though they had not
uniformly completed the requisite training. 

 Examining the
impact of those cards on the election, we first identified the
controlling inquiry: “[A]re the articles sufficiently valuable
and desirable in the eyes of the person to whom they are of-
fered, to have the potential to influence that person’s vote?”

 at 1033 (quoting Nestle Ice Cream Co. v. NLRB, 

,
583 (6th Cir. 1995)). We concluded that the promised cards
were. Id. at 1033–34. By offering employees a benefit that they
were not qualified or entitled to receive, the union promised
to gift “access to more lucrative jobs at a far lesser cost.” Id. at
1033.
   As is apparent from previous decisions, no single ap-
proach controls. Jam overstates the uniformity of Board prec-
edent and circuit cases when it argues that “[t]he Board and
the courts consistently have applied” the principle that
“granting a benefit during the critical period is, itself,
28                                                    No. 22-1122

sufficient to create an inference of objectionable conduct.” Ra-
ther, over time the Board and this court have approached the
question of objectionable preelection benefits in different
ways.
                                 B
   With that, we determine whether the Board’s articulated
framework is reasonable. It helps at this juncture to restate the
considerations at play.
    When examining elections, one question the Board asks is
whether a party has engaged in objectionable conduct—con-
duct that has a “tendency to influence” the outcome of the
election. Gulf State Canners, 242 NLRB at 1326; B&D Plastics,
302 NLRB at 245. One form of objectionable conduct is when
a union coerces employees by conferring benefits pre-elec-
tion. See Savair, 414 U.S. at 278–80; see also Warner Press, Inc. v.
NLRB, 

 (7th Cir. 1975); Jam Prods., 

. But not every pre-election grant of benefits is necessarily
coercive. See, e.g., Virginia Concrete, 338 NLRB at 1184–85;
Sequel of New Mexico, 361 NLRB at 1125–26. So, the Board’s
inquiry here, as we see it, is to identify when a conferral of
pre-election benefits is coercive and thus objectionable. On
that point, the Board’s decision states, “Where an objecting
party alleges a union granted access to a hiring-hall benefit
during the critical period, it has the burden of proving not just
that the union did so but also that the benefit was one to which
the employees were not otherwise entitled.” If that burden is
met, the Board then will draw a rebuttable inference that the
benefits granted are coercive.
   We examine that language in parts, beginning with the
Board’s guidelines for inference drawing. Jam says that under
No. 22-1122                                                    29

the Board’s framework, a union granting voters hiring-hall
benefits during the critical period is insufficient to create an
inference of objectionable conduct. Jam takes issue with that
approach, but we identify no reversible error. As precedent
confirms, neither we nor the Board necessarily infer coercion
when evaluating whether a grant of benefits during the criti-
cal period is objectionable. For example, this court in Chicago
Tribune Co. determined that a grant of apprentice cards dur-
ing the critical period was not objectionable. 943 F.2d at 797.
In reaching that conclusion, we simply examined the record
and agreed that the employer had not “presented evidence
that the Union issued apprentice cards outside of the ‘normal
course of business’ in order to influence the vote in the elec-
tion.” Id. (quoting St. Francis Fed’n of Nurses & Health Pros. v.
NLRB, 

, 850 (D.C. Cir. 1984)). The Board has like-
wise declined to always find a grant of benefits objectionable
simply because a party provided voters something of value
during the critical period. See, e.g., Topside Construction, 329
NLRB at 898; Drew Electric, 312 NLRB at 591. As such, the
Board can reasonably require Jam to show something more
than voters receiving a critical period hiring-hall benefit be-
fore inferring coercion.
    Next, we examine exactly what the Board required Jam to
show before it would infer coercion. As a prerequisite to in-
ference drawing, the Board required Jam to demonstrate not
only that Local 2 granted voters access to hiring- hall referrals,
but also that “the benefit was one to which the employees
were not otherwise entitled.” (emphasis added). This require-
ment is reasonable. The Board has previously used identical
or similar language to determine if a grant or promise of ben-
efits during the critical period is objectionable, and we see
nothing unreasonable about deploying that construction here.
30                                                   No. 22-1122

See, e.g., Mailing Services, 293 NLRB at 565; Drew Electric, 312
NLRB at 592; Topside Construction, 329 NLRB at 898; Go Ahead
North America, 357 NLRB at 78; McCarty Processors, 286 NLRB
at 703.
     This and other circuit courts have used a comparable anal-
ysis when ascertaining whether a grant of benefits is objec-
tionable. See, e.g., Chi. Tribune Co., 943 F.2d at 797 (“[B]ecause
there was no change in the status quo with regard to employee
benefits, the Union’s issuance of the apprenticeship cards did
not interfere with the election.”) (emphasis added); King Elec.,
Inc. v. NLRB, 

 (D.C. Cir. 2006) (“As we have
indicated, a union’s grant to employees of a benefit to which
they are not otherwise entitled, during an election campaign, is
still objectionable … .”) (emphasis added). The “not otherwise
entitled to” distinction makes sense. If voters are already en-
titled to receive a benefit well in advance of an election, such
a benefit could not potentially influence a vote.
    To be clear, in our prior decision in this case, we wrote “the
Board has held that a union is forbidden from providing vot-
ers anything of ‘tangible economic benefit’ during the critical
period before the election.” Jam Productions, 

.
Jam references that statement for the proposition that the
grant of any tangible economic benefit during the critical pe-
riod at least raises an inference of coercion. In contrast, the
Board understands that statement to mean that a union is
barred from providing voters “any benefit that reasonably
can be seen as an economic inducement to vote for a union.”
The quote above, read out of context, could be misinterpreted.
So, we take this opportunity to reiterate that, when examining
whether benefits improperly influenced a representation elec-
tion, the guiding question is whether “what is offered is
No. 22-1122                                                                31

‘sufficiently valuable and desirable in the eyes of the person
to whom they are offered, to have the potential to influence
that person’s vote[.]’” 

 at 1044 (quoting River City Elevator,

).
    Under that standard, we discern no unreasonable applica-
tion of law in the Board’s decision. General access to hiring-
hall referrals is not a benefit offered in the critical period if it
was available to voters on the same, routine terms all along.
As such, not every hiring-hall benefit will “taint[] the fairness
of the election,” and the Board need not automatically draw
an inference of coercion just because a voter received hiring-
hall referrals during the critical period. 

 Our prior decision
recognized this when it explained Local 2’s referrals might
amount to improper inducement “[i]f the jobs were in fact …
previously unavailable to those employees.” 

 (empha-
sis added); see also 

 (explaining that denial of an ev-
identiary hearing confined Jam’s ability “to demonstrate that
the referrals were in fact an aberration from Local No. 2’s or-
dinary referral operating system”) (emphasis added). 10 As is
apparent from that language, our prior opinion did not hold


    10 Another case Jam cites, Freund Baking, 165 F.3d at 931–32, also must

be considered in context. There, the D.C. Circuit stated, “the Board has
held that a union may not give voters anything of ‘tangible economic ben-
efit’ during the critical period before an election.” 

 (citing Mailing Ser-
vices, 293 NLRB at 565–66). That statement is broad, but the decision to
which Freund cites for that proposition undermines the rule’s scope. As
support for the above-quoted text, Freund cites to Mailing Services, where
the Board held that unions are “barred in the critical period … from con-
ferring on potential voters a financial benefit to which they would otherwise
not be entitled.” 293 NLRB at 565 (emphasis added). Given that context, the
rule of Freund is not so broad that a union provision of any benefit is coer-
cive or demands an inference of coercion.
32                                                            No. 22-1122

that any critical period referrals would amount to objectiona-
ble conduct or an adverse inference.
    At bottom, the Board reasonably interpreted and applied
the applicable law in this area. Examining the facts of Jam’s
objection, the Board correctly surmised that it had “not previ-
ously articulated how cases addressing employment- and hir-
ing hall-related benefits during a pre-election critical period
interact with [the] B&D Plastics framework for assessing other
types of grants or promises.” The Board took the opportunity
to do so in this case, and considering our deferential standard
of review, we hold that the Board used a reasonable frame-
work.
                                    III
    We now turn to the Board’s factual findings, which we will
uphold if they are supported by substantial evidence. Amer-
iCold Logistics, 

; 

(e). First, the
Board held that Jam “failed to prove the union provided re-
ferrals to the [Shaw] voters to which they were not otherwise
entitled” and, as consequence, found it unnecessary to draw
an inference of coercion. 11 Second, the Board went one step
further and determined that “even if the increase in job refer-
rals during the critical period were construed as a grant of
benefits,” the union’s explanation for the referrals rebutted




     11 We acknowledge there is room for disagreement on whether the

question of Shaw voters receiving benefits to which they were not other-
wise entitled is a factual or legal one. Either way, our holding remains the
same. To the extent this determination is legal, it is reasonable. And if the
determination is factual, substantial evidence supports it.
No. 22-1122                                                    33

“any inference of coercive, election-related timing or pur-
pose.”
                                A
    Before considering each of these findings, we first address
whether participants are “entitled” to hiring-hall work. Ac-
cording to Jam, the discretionary nature of the hiring hall
means that Shaw voters could not have received referrals they
were otherwise entitled to have. Local 2’s discretion in allo-
cating referrals, says Jam, means hiring-hall participants are
never “entitled” to receive referrals: “[T]he evidence showed
that Local 2’s referral process was wholly discretionary and
that no participants in the hiring hall, including the Shaw vot-
ers, were entitled to receive any particular referral.” For Jam,
“entitlement” requires discernible pre-existing referral prac-
tices, which it contends the union “did not and could not
identify,” thus imposing an unrealistic evidentiary burden on
Jam. Jam contends it is unreasonable to ask employers to
prove deviation from a status quo when that status quo in-
volves discretion. In response, the Board asserts “the Union
made referrals based on a variety of legitimate factors” and
highlights the Shaw voter’s entitlement to referrals as a gen-
eral matter.
    In its decision, the Board rejected Jam’s premise that the
hiring hall lacks discernible operating procedures. Specifi-
cally, the Board examined the evidentiary record—including
witness testimony on how the hiring hall functions—and con-
cluded the Shaw voter referrals “were consistent with Local
2’s usual practices.” Implicit in that conclusion is the existence
of meaningful “usual practices.” The Board also rejected
Jam’s position that lack of entitlement to specific jobs means
hiring-hall referrals are benefits which the Shaw voters were
34                                                           No. 22-1122

not “otherwise entitled” to receive. On that point, the Board
reasoned that even if the Shaw voters were not entitled to any
particular job, they were still “‘entitled’ to referrals as a gen-
eral matter, in that they were eligible to receive referrals once
they had enrolled in the system, which any stagehand (Local
2 member or otherwise) could do.” Put another way, the
Board determined that the Shaw voters were entitled to nor-
mal participation in the hiring hall. The Board defends this
analysis on appeal, contending “the Board reasonably found
that Jam failed to show the [Shaw] voters received a new ben-
efit” and arguing “the Union acted consistently with its
longstanding, customary practice” of making referrals based
on “considerations that included availability, experience, and
skills.”
    Substantial evidence supports the Board’s findings that
the hiring hall has an ordinary manner of operating, and that
Shaw voters were entitled to participate in that operation,
even if they were not “entitled” to particular jobs. First up is
whether the Board could conclude “usual practices” gov-
erned the Shaw referrals. Though the record suggests that Un-
ion personnel retained significant discretion in granting refer-
rals, the system was not devoid of common practices. 12 As ex-
plained above, Herrmann considered union membership, em-
ployer preferences, special skills, availability, and the like.
When asked, he agreed that those “criteria” “guided [him]
when [he was] … making work assignments.” Herrmann also


     12 In its decision, the Board remarked that Herrmann and Carlson’s

discretion in handing out referrals “is not legally significant.” We under-
stand that to be a comment on the different legal rules applicable to exclu-
sive and non-exclusive hiring halls. We do not interpret the comment as
the Board ignoring the role discretion plays in this analysis.
No. 22-1122                                                             35

testified that he did not know which Shaw Crew members
were eligible to vote, and explained that he did not “abandon”
his identified “criteria” for making referrals as the Jam elec-
tion approached. He further denied that “anyone else at Local
2 ever instruct[ed] [him] to make sure to refer JAM employees
to jobs in the weeks before the election.” For his part, Carlson
testified that he made no promises to Jam stagehands as to
how much work they would get in return for supporting the
union, and denied ever instructing Local 2 personnel to grant
Jam employees preferential treatment in hiring-hall referrals.
     The Board credited Herrmann and Carlson’s testimony,
explaining they “testified that the referrals made during the
focal period (and indeed during the entire critical period) fol-
lowed Local 2’s usual practice.” “[N]o direct evidence” con-
tradicted that testimony, and the Board was entitled to believe
it. The record therefore contains substantial evidence that the
Shaw referrals were made per “usual practices.” The exist-
ence of a cognizable status quo also deflates Jam’s arguments
that the Board saddled it with an unrealistic evidentiary bur-
den. 13
    We turn next to the Board’s conclusion that Shaw voters
are “otherwise entitled” to participate in the hiring hall even
if they are not entitled to particular jobs. Herrmann testified


    13 Jam also implies that the Board required direct evidence of favorit-

ism, but we do not read the decision as going that far. Instead, the Board
pointed out that witnesses testified to no special treatment, and it noted
there was “no direct evidence to the contrary.” We do not interpret that to
mean direct evidence is always necessary, even if Jam’s indirect evidence
of special treatment did not persuade the Board here. In fact, the Board
“acknowledge[d] that a proven statistical disparity could arguably
demonstrate favorable treatment or a grant of benefits.”
36                                                          No. 22-1122

that participation in the Local 2 hiring hall is open and acces-
sible. In fact, the hiring hall allowed a wide variety of individ-
uals (both union and non-union) to register, including people
who walk into the union office from off the street. And, as
Herrmann explained, any registered individual is eligible to
start receiving referrals under the hall’s normal operation.
This is shown by the fact that some Shaw voters had obtained
Local 2 referrals years before the representation election.
Amidst that backdrop, Jam identifies no reason why the Shaw
voters were not entitled to participate in the hiring hall’s or-
dinary operation. Were there a special participation require-
ment that Local 2 relaxed because of the impending election,
our conclusions might change. But on this record, substantial
evidence supports the Board’s conclusion that Shaw voters
were otherwise entitled to basic hiring hall participation. 14
What matters here is not, as Jam contends, a distinction be-
tween entitlement versus eligibility. Instead, the point is that
no evidence shows Local 2 afforded the Shaw voters special
treatment merely by allowing them to participate. Substantial
evidence supports the Shaw voters’ “‘entitle[ment]’ to refer-
rals as a general matter.”
    The Board’s reasoning here is not unprecedented, either.
In Topside Construction, the Board affirmed a decision holding
that individuals were generally “entitled” to hiring hall par-
ticipation under that hall’s ordinary operating terms. 329
NLRB at 898. The employer there accused the union of


     14 As we explained above, the Board’s requirement that Jam show vot-

ers received hiring-hall benefits to which they were not otherwise entitled
is a reasonable one. So, Jam’s failure to demonstrate that here means Local
2 was not obligated to entirely suspend referrals to the Shaw voters during
the preelection critical period.
No. 22-1122                                                    37

bypassing its standard job referral practices to benefit voters
and induce their support for the union. Id. The Board disa-
greed, finding no evidence that the union “made promises to
obtain for the employees work or priority in employment
which they were not otherwise entitled to receive from the
Union simply as employment applicants to the hiring hall.”
Id. Similarly, even though the employer “introduced evidence
that various of its former employees received referrals
through the union hiring hall,” the Board determined that
there was insufficient evidence to establish the union “fa-
vored [those] individuals over others of like standing under
the hiring hall practices.” Id. at 898–99. Thus, Topside Construc-
tion reasonably supports the Board’s holding that voters are
generally entitled to participate in the Local 2 hiring hall un-
der its ordinary rules. Absent a finding of “improprieties in
the operation of [the union] hiring hall referral process,” the
referrals were not objectionable. Id. at 899. By contrast, hiring-
hall referrals may support a finding of union improprieties
when the hall provides voters with special treatment they
were not entitled to before.
    The D.C. Circuit’s decision in King Electric, 

,
does not alter this analysis. There, a union filed an election
petition and told voters during the campaign they were eligi-
ble for hiring-hall referrals, even though the hall was not gen-
erally open to the public. 

 As the union explained it, King
Electric employees could take referrals “regardless of whether
the union won or lost the election,” because “at least 51% of
King’s employees had signed authorization cards.” 

 King
Electric objected to that promise of benefits as improper. 

. The D.C. Circuit explained “a union’s grant to employees
of a benefit to which they are not otherwise entitled, during an
election campaign, is still objectionable … whether or not
38                                                    No. 22-1122

conditioned on how employees vote in an election.” 

 (emphasis added). So, the question became whether
the union strictly applied the “51% rule,” such that the King
Electric employees were “entitled” to referrals via the signed
authorization cards, or “whether [the rule] was merely some-
thing that was in the union’s discretion to offer in appropriate
situations—perhaps when necessary in order to encourage
pro-union votes.” 

. The court held for the employer,
reasoning “the hearing officer never made a finding that the
union treated the ‘51% rule’ as binding on it and thereby un-
conditionally available to employees.” 

    Jam reads King Electric to mean that “where a union re-
tained discretion whether to award … benefits, the recipients
were ‘not actually entitled’ to the referrals.” Our takeaway is
different. As we see it, the animating question in King Electric
was whether employees were entitled to access the hiring hall.

 (“According to the Board, without the ‘51% rule,’
King employees would not have been eligible for … refer-
rals … .”). If application of the “51% rule” was discretionary,
then the union could be seen as giving the employees special,
undeserved access. 

. But if the rule was “invariably
employed,” then perhaps the employees were already enti-
tled to participate. 

 Thus, the dispute in King Electric sets it
apart from this case. Here, the question is not whether Shaw
voters were barred from the hiring hall—they were not. The
issue is how referrals were made within the hall. On that
question, King Electric offers no insight. Therefore, we do not
disturb the Board’s conclusion that Shaw voters were “enti-
tled” to referrals as a general matter.
No. 22-1122                                                   39

                               B
    We turn next to Jam’s main contention—that Local 2 sin-
gled out Shaw voters for special, undeserved treatment be-
yond usual hiring hall practices. For support Jam points to a
collection of statistical trends. It emphasizes that “the raw
number of referrals to the Shaw voters increased sharply at
the beginning of the focal period”; “Shaw voters began to re-
ceive union referrals immediately after the petition for elec-
tion was filed”; more Shaw voters began accepting referrals
as the election approached; and the rate of referral to Shaw
voters rose during the focal period as well. Jam believes these
trends “show[] not only that the raw numbers of lucrative re-
ferrals to the Shaw voters increased dramatically in the critical
and focal periods, but also that the Shaw voters were targeted
and favored with disproportionate numbers of referrals.” The
Board was not persuaded and defends its weighing of the ev-
idence on appeal.
    Substantial evidence supports the Board’s decision that
the Shaw voters were not given special treatment. To start, the
evidentiary hearing yielded witness testimony in addition to
hiring-hall data. The Board was entitled to credit testimony
from union personnel explaining that they did not grant spe-
cial treatment to Shaw voters. Herrmann testified that he used
personal discretion when assigning referrals but consulted a
set of general factors. For instance, he would always try to sat-
isfy employer requests, provide specially skilled workers
when necessary, and accommodate stagehand schedules. He
further testified that he did not let his personal feelings to-
wards hiring hall participants cloud his work and that he did
not abandon his usual criteria when the Jam election ap-
proached. Carlson, who occasionally assisted with assigning
40                                                  No. 22-1122

referrals during the relevant period, likewise testified that the
Shaw voters received no special treatment. He said he did not
promise voters access to work in exchange for union support
nor did he consider a participant’s status as a Jam employee
when making referrals. He averred he never ordered the
Shaw voters to receive priority treatment in the hiring hall.
Different factfinders may weigh that testimony differently,
but the Board is allowed to credit it.
    Jam tries to refute this testimony through statistical evi-
dence, but the data trends are not so incontrovertible that the
Board had to accept Jam’s version of events. The statistics Jam
points to demonstrate that Shaw voters received increased re-
ferrals as the election approached and generally suggest that
the eligible voters received more referrals than other identi-
fied groups during the focal period. True enough, but those
trends do not necessarily mean Local 2 was giving Shaw vot-
ers special treatment. As the Board points out, the full picture
is more complex, with the data suffering from certain weak-
nesses.
    For example, Shaw voters who received their first referrals
between September 2015 and May 2016 had a higher referral
rate during the focal period than other non-voter stagehands
who also received their first referrals at that time. But the rec-
ord does not identify how many of those other non-voter
stagehand “comparators” were still available or looking for
work during the focal period. Some of those individuals
might have accepted referrals in the fall of 2015 and then
moved away or found other work, skewing the referral rate
data. The record also contains no information on how many
referrals the “comparators” received prior to the focal period,
thus failing to show change over time. We highlight these
No. 22-1122                                                              41

open questions not to reweigh the evidence, but to illustrate
the tangle of factual issues facing the Board and thus the room
for reasonable disagreement on which conclusions to draw
from the evidence.
    Other data sets, many of which the Board emphasizes on
appeal, also support its decision. Post-election data demon-
strates that referrals to Shaw voters rose sharply over the
summer of 2016, when the union would have no incentive to
give them special treatment. And referrals to non-voting
Shaw Crew members increased “by an even greater percent-
age during the focal period” compared to the Shaw voters.
The Board’s Table A 15 is also instructive, as it traces the
general referral patterns for different groups across time. Ex-
amination of this table suggests that Shaw voters received in-
creased referrals largely in lockstep with busy time periods.
Thus, the Board could conclude that the increased referrals
had more to do with seasonal needs than the representation
election.
     Finally, the Board did not have to infer that the union fa-
vored Shaw voters from the referral patterns Jam identified.
Jam points out that the union maintained a list of eligible
Shaw voters and that those voters were referred to jobs as a
group with some frequency. According to Jam, these facts in
combination require an inference that the union kept the voter
list to “target and favor the Shaw voters or otherwise support
the existence of such favoritism.” But without more, the mere
fact of group referral does not demand an inference of favor-
itism. Jam did not examine witnesses about the grouping


    15 Table A is available in the text of the Board’s decision. Jam Prods.,

Ltd. v. NLRB, 

, 13-RC-160240, at 7 (Sept. 30, 2021).
42                                                No. 22-1122

patterns, and the Board did not need to draw the inference
Jam invites. The same is true for the fact that Justin Huffman,
a Shaw voter and the union’s observer during the election, re-
ceived the most referrals out of the Shaw Crew. For Jam, the
“reasonable inference is that Huffman was favored most be-
cause he was an active Union supporter.” While that may be
one possible inference, it is not the only one. Herrmann testi-
fied under oath that during the focal period he was not aware
Huffman was a main Local 2 contact amongst the Shaw Crew.
So, a different, reasonable inference is that Herrmann did not
deliberately favor Huffman as a reward for his organizing ef-
forts. We recognize how the data looks, and we understand
that different factfinders might weigh the evidence differently
and draw different conclusions. But our review is not de
novo, and we do not reweigh evidence. ADT, 54 F.4th at 97.
The Board did not commit reversible error when it credited
Herrmann’s testimony and declined to draw conclusions fa-
vorable to Jam.
    Jam urges us to force the Board to make certain inferences.
Under our deferential standard of review, we decline to do so.
The Board concluded that the union did not give special treat-
ment to Shaw voters, and it relied on substantial evidence in
so deciding. The Board carefully examined the statistical evi-
dence and witness testimony, concluding that the union did
not engage in election misconduct. We do not disturb that de-
cision.
                              C
    The Board decision took a “belt and suspenders” ap-
proach, finding that—even if the hiring-hall referrals during
the critical period raised an inference of coercion—the union’s
explanations rebutted that inference. We thus turn our
No. 22-1122                                                   43

attention to Local 2’s proffered explanations for why Shaw
voters received increased referrals as the election approached.
Jam criticizes Local 2’s explanations and the Board’s
treatment of them, arguing the Board held “Local 2 to a non-
existent rebuttal burden and accept[ed] Local 2’s factually un-
supported ‘explanation’ of its conduct as sufficient, despite
substantial evidence to the contrary in the record as a whole.”
    The union’s first explanation for its referral patterns cen-
ters around the April 2016 focal period. Per Local 2, the in-
crease in referrals to Shaw voters during the focal period is
attributable to the hiring hall being extraordinarily busy at
that time. Indeed, the Board argues that the data shows Shaw
voters “received limited referrals during the coldest months,
followed by a marked increase in the spring.” The Board de-
cision credited that explanation, and it relied on substantial
evidence in doing so. Witnesses confirm that the spring of
2016 was the busiest time in the hiring hall’s history to date.
They testified further that the demand for labor fluctuates sea-
sonally, with the need for work starting to rise at “the end of
March, beginning of April, until the end of September, begin-
ning of October.” The focal period coincided with spring,
meaning a general increase in across-the-board referrals
could be expected. Table A also strongly suggests that the fo-
cal period was exceptionally busy, as it shows referrals in-
creasing to all groups at that time. So, even if certain isolated
days during the focal period were busier than others, the
Board still relied on substantial evidence in analyzing general
correlations across the entire focal period.
    Local 2’s second explanation relates to referrals granted
during the abeyance period before the focal period. According
to the union, referrals increased during that time because Jam
44                                                 No. 22-1122

had just collectively terminated the entire Shaw Crew, and
those individuals were looking for work. The increase in re-
ferrals to Shaw voters in 2015 indeed took place after Jam ter-
minated the Shaw Crew, so the Board was entitled to believe
the union’s offered explanation.
    One final point: Per Jam, the union also gave Shaw voters
preferential treatment by “failing to apply its standard drug-
testing requirements to the Shaw voters.” On this count, the
Board found that Jam failed to present sufficient evidence of
a drug test policy and violation thereof. The Board concluded
that “the record contains no evidence regarding whether new
enrollees were tested during the relevant time period.” Fur-
ther, the rules relevant to the drug testing requirement were
not submitted into evidence. Both findings draw support
from substantial record evidence. Herrmann and Carlson
both testified that they did not know whether the Shaw voters
enrolled during the critical period underwent drug testing.
The two stagehand witnesses testified that they were not per-
sonally drug tested, but they started receiving referrals before
the critical period began. Jam presented no other evidence
suggesting that Local 2 waived drug testing requirements for
the Shaw voters. Thus, the Board relied on substantial evi-
dence in rejecting Jam’s drug testing arguments.
                        *      *      *
   The Board decision reasonably applies substantive labor
law, and substantial evidence supports its factual findings. So,
we GRANT the Board’s application for enforcement of its Jan-
uary 11, 2022, Order in Board Case No. 13-CA-284761.