Would you hire an applicant without conducting an interview? While
almost no one would say "yes" to this question, many people probably
have hired applicants without conducting a reference check. Like any
other selection tool, the reference check is not fool-proof. But, as I
will discuss below, a reference check is generally worth the time it
takes to perform one.
References have a number of legal liabilities associated with them,
which is probably one of the main reasons that companies are cautious
in discussing a past employee. Most of these legal liabilities involve
the reference provider, although there are some legal issues for the
reference checker as well. The focus here is on the reference checker,
so I will spend relatively little time discussing legal issues.
It should be noted that there are different types of reference
providers. Many job applicants will offer the names of people who
know them (i.e., personal references). These individuals may be former
supervisors, peers from earlier jobs, and perhaps even friends. Be
aware that they may not necessarily be the people they say they are!
Most company application forms also require the applicant to include
the names of past supervisors. Although those names are not always
accurate, past supervisors may be more reliable about sharing
information than personal references.
There are probably two major reasons a prospective employer should
conduct reference checks. First, a reference check enables the
employer to verify and corroborate information provided by the
applicant. Because a relatively large number of applicants provide
false information, it is important to verify the information that
applicants provide. One of the few ways to detect false information is
by confirming the information (e.g., dates of previous jobs) with the
former employer.
A second reason to conduct a reference check is to reduce liability
associated with a negligent hiring situation. Briefly stated, a
negligent hiring suit is based on the hiring of an applicant that the
employer knows, or should have known, possesses some trait, or has a
history of prior conduct, which would create an undue risk or harm to
others in carrying out his or her employment responsibilities. Even if
a thorough and proper reference check does not reveal a potential
problem, by doing the reference check, the employer has taken a step
toward documenting that "due diligence," the opposite of negligence,
was used in the hiring process.
Probably the major reason that companies skip doing a reference
check is that they question the accuracy of the information they will
receive. Despite the perception that references often have absolutely
no validity, a recent study showed the correlation between the
reference information and job performance is about .18, but corrected
for statistical artifacts, it is about .29 (correlations range between
.00 to 1.00, with most good selection tools somewhere between
.30-.50). Although .29 is still just a modest correlation, and other
selection tools (e.g., structured interviews) have more impressive
correlations, a reference check may get at attributes that cannot be
easily assessed elsewhere.
Why is the correlation quite modest for reference checking? I
believe that the main reason is that reference providers tend to be
quite lenient in the information they provide and generally speaking,
are reluctant to say something that is terribly negative about a
former employee. This hesitancy is probably due to at least two
factors: fear of legal repercussions (e.g., a lawsuit will be filed
for defamation) and guilt associated with hurting a past employee's
chances for employment. Based on these issues, the likelihood of a
false negative (i.e., a problem applicant being labeled "acceptable"
based on an overly lenient reference) is relatively high. Of course,
there are also stories indicating that false positives (i.e., a good
applicant being labeled "unacceptable" based on an overly severe
reference) sometimes occur, such as when a past supervisor has a
personal vendetta against the employee. In some instances, all the
reference checker may get is a confirmation (or, possibly, a
disconfirmation) of an applicant's dates of employment and job
title. Again, however, this may be useful information in that it might
reveal some discrepancies with what the applicant has provided (e.g.,
information listed on his or her resume). The bottom line is that
while reference checking is far from a perfect science, it may reveal
valuable information.
Overall, there are several good reasons for doing, and few reasons
to not do, reference checks. Each organization should develop policies
regarding the inclusion of reference checks and how they are
conducted. Below are some suggestions for companies that choose to
perform reference checks.
Do: Standardize the reference checking process as much as
possible; work with a standard set of questions.
Do: Make sure that your questions are job-related (e.g.,
cover relevant skills).
Do: Try to reach past supervisors; they are more likely to
give accurate information than personal references.
Don't: Ask questions that are illegal to ask the applicant
(e.g., about disabilities).
Don't: Fail to document the information you receive, even if
it is a statement from the past employer that they don't provide
reference information.
Don't: Do a reference check for some applicants, but skip
the process for other applicants. This might be viewed as an
indication of discrimination.