Neutrality agreements change the landscape
of union organizing. A study by
Eaton and Kriesky conducted in the late
1990s examined 170 campaigns where the
employer and the union agreed to neutrality
with a card-check provision. With the
aid of such agreements, unions prevailed
78.2 percent of the time.19 Comparing this
result to the 46 percent union success rate
in contested NLRB elections, the authors
concluded that while neutrality arrangements
improve a union’s chances, the
results are not nearly as stark as those in
the public sector where unions won 85
percent of secret-ballot elections in 1995.
We believe that the authors dismissed
their results too quickly. We say this because
of unions’ policy of petitioning for an election
only when they are certain of having
more than 50 percent support of the unit’s
members. This implies that in the 54 percent
of the elections that unions lost, the
unions went into the election with majority
support, but did not get all of those
votes. This means that if there had been a
card check instead of a ballot, those unions
would likely have been recognized.
The following example contrasts the
effectiveness of neutrality agreements for
organizing with that of contested elections.
Say that a union is targeting workers
at two hundred different employers. One
hundred of the employers sign a neutrality
agreement with a card check, while the
NOVEMBER 2006 Cornell Hotel and Restaurant Administration Quarterly 341
THE HOTEL INDUSTRY’S SUMMER OF 2006 CHR LEGAL ROUNDTABLE
CQ292810.qxd 10/4/2006 10:08 PM Page 341
other one hundred do not do so. Applying
Eaton and Kriesky’s results for this example,
the union would organize seventy-eight of
the “neutral” employers. The union then
attempts to organize the other one hundred
employers using secret ballots instead of
neutrality agreements. Of these one hundred
companies, the employees in twenty
companies do not want a union. Because
fewer than 50 percent of those employees
sign authorization cards, the union walks
away. The union gets more than 50 percent
of the employees to sign cards at the
remaining eighty companies. Note that
under a card-check agreement, the matter
would end there and all eighty of these
companies would be unionized. Because
there is an election, however, the employer
tells its story (or intimidates the employees)
during the campaign period. Again
applying Eaton and Kriesky’s percentages,
the union prevails in thirty-seven elections
(46 percent of the elections). A casual
reading of this example would conclude
that neutrality with a card check yielded a
78 percent union recognition and election
yielded just under a 50 percent union success
rate. That conclusion, however, omits
the effects of the twenty cases where the
union did not even attempt an election.
Thus, the result when employers declined
a neutrality agreement and insisted on
elections was just under 40 percent union
success, given that twenty companies are
not in the election data set. What this
means for managers is that when there is
enough employee interest to warrant an
election, (1) the company’s chances of
becoming unionized are less than 50 percent
under the NLRB’s election procedures,
but (2) a unionized bargaining unit
is nearly guaranteed under a neutrality
agreement with a card-check provision.
We are certain that employers are aware
of the logic found in the above example.
Why, then, would an employer ever
accede to a neutrality agreement in the
first place? As we explain next, the answer
is external factors.