Worker Misclassification, Co-Employment, and the Efficacy of
Assignment Limits--How Not to Solve the “Microsoft”
Problem
By Alden J. Bianchi,
Esq.
Mintz, Levin, Cohn, Ferris, Glovsky and Popeo, P.C., Boston, MA
“There is always an easy solution to every human
problem--neat, plausible, and wrong.” (H.L. Mencken)
Employers routinely rely upon a broad range of
“contingent” employment arrangements to manage cyclical
demands for their products and services, supplement their core
workforce, and assist with new product roll-outs, among other things.
While the term “contingent” worker can be applied broadly
to include all manner of leased employees, freelancers, on-call
employees, contract/technical workers, and part-time, seasonal and
temporary employees, it is most often encountered in connection in
with the placement of workers by staffing firms at the workplaces of
their clients.
With staffing firms employing 2.96 million U.S. workers each day
during 2006,1 the legal and
regulatory structures governing contingent workers are critically
important, and, one would hope, straightforward. But this is not the
case. The use of contingent workers arose over the last few decades in
response to the demands of a global economy with its relentless
competitive demands. The laws governing contingent workforce
arrangements in the United States, however, are still based largely
upon common law notions of master and servant. The failure of legal
and regulatory structures to keep pace with this growing complexity
imposes upon staffing firms and client companies an added layer of
compliance-related risk, as they endeavor to comply with a broad range
of federal and state labor, employment, tax, and benefits laws.
A central issue in the regulation of contingent workers involves
the proper classification of workers--is the worker an employee of the
staffing firm, the client company, or some other entity (e.g., an
independent contractor). For the purposes of many federal and state
employment laws (and for withholding and employment tax issues),
workers may be employees of more than one entity. That is, they may be
“co-employees.” But where the regulation of tax-qualified
retirement plans and most welfare and fringe benefit plans is
concerned, the concept of co-employment is generally not
recognized. For these purposes, employment status is instead based on
the application of a “multi-factor” test set out in
Nationwide Mut. Ins. Co. v. Darden, 503 U.S. 318 (1992), that
measures the extent to which the employer controls both the ends to be
achieved and the means of achievement and establishes the identity of
the employer and the employee, in each instance, based upon the
outcome of that test.
The disparate approach of the labor laws, on the one hand, and the
tax and benefits rules, on the other, ultimately led to unfortunate
consequences for Microsoft Corporation in a series of trial and
appellate cases in the late 1990s. These cases, which collectively
came to be referred to in benefits circles as simply “the
Microsoft case,” began with the settlement of an IRS employment
tax audit that resulted in the reclassification of a group of
“freelancers.” See, e.g., Vizcaino v. Microsoft
Corp., 120 F.3d 1006 (9th Cir. 1997). Both before and after the
IRS employment tax audit, the freelancers were not covered under
Microsoft's benefit plans. As part of a settlement with the IRS, many
of these freelancers continued to work for Microsoft, but were
transferred to staffing firms.
Following the IRS settlement, the freelancers who were placed with
staffing firms sued Microsoft claiming that they were really common
law employees of Microsoft, and that, as such, they were entitled to
certain retirement, welfare, and employee stock purchase rights. The
essential message--if not the exact holding--of Microsoft is
that, workers hired through a staffing firm who are intentionally
excluded from the service recipient's benefit plan coverage may
nevertheless have rights under pension, welfare, fringe benefit, and
stock-based programs of the service recipient, if they are determined
to be the recipient's common law employees (based on application of
the multi-factor test) and are not otherwise expressly and properly
excluded from plan participation under the express terms of the plan.
NOTE: The extent to which classes of employees (such as “workers hired
through staffing companies”) may be excluded as a matter of plan
design varies depending on the benefit involved. Classes of employees
can, for the most part, be included or excluded under tax-qualified
retirement plans, but, as explained below, the exclusion of some
workers who are common law employees can affect non-discrimination
testing. While more liberal in this regard, similar constraints affect
many but not all welfare and fringe benefit plans. While non-qualified
stock options can include or exclude participants at will (and can
cover employees and non-employees), employee stock purchase plans
(under Code §423) may not.
Microsoft Corporation finally settled with the freelancers in 2001
for just shy of $100,000,000. The sheer size of the settlement quickly
got the attention of the staffing industry and their client companies.
Microsoft came to stand for the proposition that there is a
certain benefits-related risk, which came to be referred to as
“Microsoft” or “co-employment” risk, that goes
hand-in-hand with the use of workers hired through staffing firms.
Shortly after the Microsoft case settled, there surfaced an
urban myth that co-employment risk can be avoided by limiting the
length of staffing firm assignments. (This practice is commonly
referred to as “assignment limits” or “temp tenure
limits”). This is near complete nonsense. Assignment limits are
ineffectual for this purpose for at least two reasons: first, the
length of a worker's tenure is only one out of a long list of factors
comprising the multi-factor test, and it is unlikely to be
determinative unless in a close case; and, second, while assignment
limits may be used to avoid “leased employee” status under
Code §414(n), the leased employee rules do not change the
application of outcome of the multi-factor test. Rather, it speaks to
an entirely different set of concerns relating to non-discrimination
testing in tax-qualified retirement plans.
By way of example, assume worker A is placed by staffing firm B
with Company C, and that Company C's 401(k) profit sharing plan covers
all common law employees, as of the first day of the month following
their date of hire under the “elapsed time” method.
Company C has a twelve-month assignment limit policy for employees
hired through staffing firms. Whether A is a common law employee of
Company C is determined under the multi-factor test. That A's tenure
is limited is one factor indicating that A is not a common law
employee. But if A shows up at C's place of business on C's schedule,
works along side C's regular employees, works full-time on projects
relating to C's core business, and has no other job, etc., she could
easily be C's common law employee with rights under Company's C's
plan. C's assignment limit policy in this instance would not change
this result.
To be sure, the issues that first surfaced in the Microsoft
case raise a series of benefits-related concerns that affect all
manner of welfare, pension, and fringe benefit plans. Contingent
workers who have been denied plan participation in the absence of a
specific exclusion may well have rights under ERISA to sue for
benefits and other equitable relief. Such a plan may suffer a
“qualification failure,” which could lead to the loss of
its tax-qualified status. Conversely, a tax-qualified plan covering
non-employees violates the exclusive benefit rule, which too can
result in a qualification failure. The solution is not assignment
limits, however; it lies, rather, with thoughtful plan design and
proper drafting.
Cases subsequent to Microsoft illustrate that employers are
not helpless in the matter of worker misclassification and that there
are steps that they can take to prevent worker misclassification
exposure. Among others, Bronk v. Mountain States Telephone &
Telegraph, Inc., 140 F.3d 1335 (10th Cir. 1998); Burrey v.
Pacific Gas & Elec. Co., 159 F3d 388 (9th Cir. 1998), and
Wolf v. Coca-Cola Co., 2000 WL 33164 (11th Cir. 2000)
collectively stand for the proposition that an employer may exclude
from plan participation individuals hired through staffing firms.
These individuals might include, for example, individuals who
performed technical assistance, design, research, and drafting work
for an employer's engineering department, who are not paid by the
employer itself. Plans can also include language expressly limiting
the rights of re-classified workers (so-called “Microsoft
inoculation language”). But these measures require a degree of
attention to plan drafting and maintenance that is too often lacking,
even among sophisticated employers.
There are instances where assignment limits make sense--but not for
the purpose of reducing Microsoft or co-employment exposure.
Code §414(n) treats a “leased employee” as an
employee of the client company for purposes of the Code's general
non-discrimination rules, and the related participation, coverage, and
related rules that ban discrimination in benefits or contributions
under tax-qualified plans, among others. A “leased
employee” for this purpose means an individual who is employed
by a “leasing organization,” but performs services for a
client company, provided: (i) the services are performed under the
primary direction or control of the recipient; (ii) the individual has
performed services for the lessee (or related persons) on a
substantially full-time basis for at least one year; and (iii)
the services are performed under an agreement between the leasing
organization and the person receiving the services. Nothing in Code
§414(n) requires a plan to cover leased employees, but it does
require that they be counted for non-discrimination testing
purposes.
As a result of the mechanics of the Code §414(n) “leased
employee” rules, an employer has some control over which workers
are treated as leased employees for non-discrimination testing
purposes by the judicious use of assignment limits. This would
ordinarily be an issue for an employer that relies on, but wants to
exclude from plan participation, a large cohort of staffing firm
employees relative to its regular headcount.
The solution to the “Microsoft” problem is not to
reflexively impose assignment limits. It lies instead in careful plan
design and drafting that is informed by cases like Bronk,
Burry, and Wolf, among others. These cases provide a
road map that is both reliable and easy to follow.
For more information, in the Tax Management Portfolios, see
Bianchi, 399 T.M., Employee Benefits for the Contingent Workforce,
and in Tax Practice Series, see ¶5430, Employees and
Independent Contractors.
1
American Staffing Association, May 2007 Annual Economic Analysis, http://www.americanstaffing.net/statistics/economic.cfm (site last visited December 12, 2007).
|
