The Myths and Realities of On-line Criminal History Databases
By
Frederick G. Giles, CPP
(Article first published in Loss Prevention Magazine January-February 2002)
The cornucopia of information that is the Internet has given rise to the
expectation that somewhere there must be a magical mystery database of
information on whatever and whomever we need.
There is a popular perception that there are more and more databases
threatening our individual privacy a "permanent record" of criminal history
information in which all of our lives is documented.
It must be so because it makes so much sense.
But, it's not so. There is no Lucy in the sky with data.
Our nation's courts are among the oldest institutions in this nation of
laws, and their many distinct, disjointed, and undesirable systems are part
of that heritage.
So what is the state of criminal history information in this country?
How does a loss prevention or human resources professional designing a
background screening program develop an appropriate strategy for criminal
history searches?
To
explore the first question, we will take a look how criminal history
information is housed in this country, and how misimpressions become policy.
1. The FBI National Database
The most famous and misunderstood non-source is the FBI database, the
National Crime Information Center (NCIC).
Many security professionals, having come from local, state, or federal law
enforcement, call upon friends and colleagues who still have NCIC terminal
access when they need criminal history information. Sometimes, this may be a
display of insider power intended to awe those who do not have a golden
Rolodex of contacts.
But most times, it is simply a federal crime.
FBI audit and compliance programs are getting better and better with the
government stepping up prosecution of offenders.
By
law and by compact with the states who contribute information, NCIC data is
not available without permissible purpose and legislative authority, such as
has been granted to banks and securities dealers.
It is usually not an option for the ethical and law abiding professional.
Even for those who are currently sworn law enforcement officers, improper
use of NCIC will lead to prosecution.
Recently, an elected sheriff was charged with a felony for using NCIC to do
background checks on his political opponents and their supporters.
But don't be concerned about lack of access. Even for those with legislative
authority, the use of this data is rife with problems.
The NCIC was originally intended to serve as a law enforcement source of
intelligence. It does that well and is a credit to those who maintain it.
Reporting of records to the NCIC, however, is dependent on the states or
other courts forwarding the data.
It is often
not current (as much as two years behind for some states),
and rarely has current disposition information.
Generally, an NCIC search is a beginning but not an end. NCIC users
understand that, but many employers do not.
2. State-by-State Databases
So
if NCIC is not an option, why not turn to the individual states who supply
the national database?
Unfortunately, there are issues with these repositories as well.
In
many states, the state criminal history repository was, again, designed as a
law enforcement intelligence tool.
The data housed is often late, incomplete, or simply not reported from the
originating court.
To
illustrate this point, consider the experience of one Florida retailer who
relied solely on the Florida Department of Law Enforcement state repository
for criminal history checks.
The company hired an employee who had no record in the state repository.
Unfortunately, this employee subsequently assaulted a customer.
The customer's attorney placed one phone call to the court in the employee's
county of residence that revealed a significant history of assault.
The general rule of thumb for most negligent hiring cases is "if you could
have known, you should have known."
This retailer learned an expensive lesson about the deficiencies in state
repositories.
I
have seen similar examples in most state repositories.
Having said that, Florida recently passed legislation to provide some
protections when an employer incorporates a state repository search in its
hiring program.
(See the accompanying sidebar.)
=Sidebar==================================
3. Florida's Negligent Hiring Protection
Having
used Florida Department of Law Enforcement as an example of the poor record
quality of state repositories in general, it is somewhat ironic to point out
that Florida has passed a law (Florida code 768.096) that protects employers
from negligent hiring lawsuits if the employer conducts criminal records
checks through the Florida Department of Law Enforcement along with
employment history verifications:
"In a civil action for the death of, or injury or damage to, a third person
caused by the intentional tort of an employee, such employee's employer is
presumed not to have been negligent in hiring such employee if, before
hiring the employee, the employer conducted a background investigation of
the prospective employee and the investigation did not reveal any
information that reasonably demonstrated the unsuitability of the
prospective employee for the particular work to be performed or for the
employment in general."
This was well intentioned, but perhaps not well advised, legislation.
That said, if mitigation of civil risk is one of the drivers for a criminal
history check program,
=End of sidebar ==============================
Let's run some numbers on state criminal history searches. In the fifty
states plus the District of Columbia there are forty-two central criminal
record repositories.
Of
these forty-two, half of the states want the request accompanied by a signed
release.
Eight of those twenty-one states, Alabama, Minnesota, New Hampshire, New
Mexico, Ohio, Pennsylvania, Virginia, and Wyoming, require a special
notarized release.
Four states, Delaware, Nevada, Ohio, and West Virginia, require a signed
release plus a completed ten-digit fingerprint card or a thumbprint.
These requirements add significantly to the cost, time, and logistical
complexity of accessing records.
Of
the remaining twenty-one states, only eighteen have an average turn around
time of four days or less.
Some states will only accept requests by mail and average two weeks turn
around time.
Some have stretched out as long as sixty days either due to backlogs or a
pre-notification requirement of the consumer before any records will be
reported out.
In
addition to the timelessness of the reporting, what about other costs? State
repository fees range up to $25 per name, averaging $15 paid directly to the
state agency.
This, of course, does not include any handling fee for your public record
provider or the cost of fingerprint collection when required.
Fees are subject to change without notice and are expected to increase along
with all other costs.
For the moment, let's assume that time, money, and logistical obstacles are
not an issue. Just how reliable is the state repository data, anyway?
Although the potential for statewide coverage is attractive, as is the fact
that both felonies and misdemeanors may be included, experience indicates
that the quality of the reports from these agencies is substantially lower
than for records obtained directly from the originating courts.
Remember, the state repository data is second-generation information
gathered from courts and police departments statewide, with little
consistency in format or language.
There are sometimes problems in matching arrest data with disposition data
since they come from different sources at different times.
Even in those states where there is a statutory requirement for the various
criminal justice agencies to report records to the state repository,
we
have found both significant reporting delays and in many cases known records
having never been reported to the repository at all!
An
administrator with one state repository who asked not to be identified
stated that he estimated that only about 40 percent of all criminal data
from the police and criminal courts was actually ever reported to the state
criminal record center.
Further, he thought this was pretty typical compared to other
states!
In
1993, the Defense Personnel Security Research Center (PERSEREC) published a
white paper authored by Richards Heuer entitled
"Crime andSecurity Risk."
The study was commissioned by the Director of the CIA and was one in a
series evaluating various criteria for adjudicating security clearances.
This paper evaluated the condition of criminal records in the U.S. and
determined that both the NCIC and various state repositories were flawed by
reporting problems and that the only search exercising appropriate due
diligence was one done at the originating courts.
Unfortunately, no updated study has been published. But based on current
experience, I believe a similar study today would yield the same conclusion.
With all these attendant problems, is there a place for a statewide search
in the pre-employment background process?
Perhaps. A statewide search is to a court of origin search as a shotgun is
to a rifle. A shotgun will disperse lots of little pellets covering a wider
area on a map, but will also leave a lot of target area un-hit.
With a rifle, on the other hand, one can be reasonably assured of hitting
the target exactly.
If
having the widest possible coverage is an objective, while recognizing all
the holes in coverage, a state repository search achieves that objective.
4. Courts of Origin
Always mindful of the holes inherent in state repository searches, I would
recommend that, as a minimum, an employer consider adding a search of the
county-of-residence superior court to ensure that those holes do not create
major liabilities in states other than Florida.
If
the objective of any background search is more than merely "papering a
file," searches in those courts of origin are a critical element
What are courts of origin?
5.
County Superior Court.
The most commonly searched state court is the county superior or
felony-level court, which is usually located in the county seat.
This court of origin contains the records for the felony-level cases that
originate in that jurisdiction.
A complete county felony search at that location will reveal
the following:
-Felony charges that resulted in a felony conviction,
-Felony charges that were pleaded down to a misdemeanor
conviction,
-Bench warrants issued by that court, and
-Cases that are still pending final action for all cases and
convictions that have not been sealed or expunged.
Every rule has exceptions. There are jurisdictions in which the felony level
court is actually an independent city court,
or
in which there are split divisions requiring at least two searches at
different locations to complete a countywide felony search.
Police officers are trained to "hang from the highest branch on the tree,"
meaning that if there is any discretion regarding whether an offense is
classified as a felony or misdemeanor, always arrest for the most serious
felony and let the courts sort matters out.
Not only does this practice result in many more cases being found in the
felony courts, but it also helps unclog our courts system
by
affording reasonable grounds for plea-bargaining to avoid lengthy and
expensive trials.
Many felony cases, therefore, end up as misdemeanor convictions.
6. Municipal Courts.
Misdemeanors are a source of confusion for many.
Misdemeanors are defined in different ways by different
states, but to simplify somewhat, they are generally
those offenses that can result in fines of less than $1000 or
jail terms of less than one year in a county facility.
The courts that hear misdemeanors are generally municipal courts, and they
do not always share space, indices, or administration with the courts that
adjudicate felonies.
This means that there is often not a one-stop shop for a county-wide search
for all felonies and all misdemeanors for an applicant.
There are three conditions under which misdemeanors will be revealed.
First, a search of the county felony courts will also reveal, as noted
above, those misdemeanor convictions resulting from an original felony
charge. (Note that is some courts, these are the only misdemeanors that will
be revealed).
Second, there are many courts that are unified to the extent that either all
the case activity can be searched in one index, or at the least, there is an
index for all misdemeanor courts in that county co-located in the same
county seat courthouse with the felony court.
Third is a circumstance very similar to the second, but only the
misdemeanors from that municipal court serving the county seat can be
researched at that court.
What all this means is that the requestor, who truly wants or needs
misdemeanors for a particular municipal jurisdiction, should specify that
jurisdiction in the request, or the search will be done only at the county
seat, with the possible range of results above.
In
the case of misdemeanor courts, the confusion is not only with the
requestor.
Many of the public record access difficulties come in these courts.
As
in all public sector jobs, tax cuts combined with population growth have
forced these courts to do more with less.
Looking for ways to ease that burden, periodically one of these courts will
devise new and onerous or restrictive policies to reduce their research.
The best examples of the worst are found in New Jersey, pockets of Kentucky,
and California, which flares up occasionally. In New Jersey, many municipal
courts simple refuse to accept research requests in person or at all.
=Sidebar===================================
If you don't like the way I run my court
Some years ago, one of my research agents was summarily kicked out of a
Kentucky court. Fearing that he was somehow offensive
or
in the wrong, I called up the clerk of court to discuss the issue. After
reviewing the incident with the clerk, he responded that his
court was simply too busy to respond to "all these pre-employment requests"
and that he simply was not going to do it anymore.
If
I needed that information, I could visit the state repository. I explained
my concerns about state repositories and asked if there wasn't a
compromise solution, as these were, after all, public records. At that
point, he lost his patience and said, "
Looky here. If you don't like the way I run my court, here's what you do.
First, you move yourself to this county and establish a
residence. Then, you run for this office. Then, if you are fortunate enough
to get yourself elected," raising his voice he went on,
"you can run this court anyway you want. Until then, it is my court and I
will run it the way I see fit."
That was the end of the conversation. I suspect that I was not his first
caller that day.
=End of sidebar ===============================
7. Federal Courts.
The next most common type of criminal court of origin is the federal
district court.
These courts hear cases involving violations of federal law. There may be
only one to a state, or there may be several. These courts can be searched
in person or by using a remote access. But, however they are searched, they
represent a problem for the user.
Since these courts rely very heavily on fingerprint identification, it is
very difficult to get anything other than a name and possible address match
with tools other than fingerprints.
That said, there is no substitute for a federal district search if that
search is required for a reasonable due diligence to fill a position.
Also, this search is a requirement for those employers authorized to use
fingerprints, such as registered securities dealers, as the fingerprint
search against the NCIC will not contain the most up to date adjudication
information for a given case.
For the employer with a tighter budget, a federal district search for lower
level positions, in the absence of a legal requirement or specific risk
avoidance objective, may be overkill.
While it will reveal charges not found in any state search, it is likely
that the applicant will be no stranger to the state courts.
8. Court Research.
For each of the sources of criminal history noted above, there is no single
access method. For most, a research agent is still required to physically
visit the court unless, of course, you can afford the average three-to
four-week turn around via mailed requests.
Once at the court, research is performed a variety of ways. In some courts,
a local public access terminal is available that can provide not only index
information, but also complete case information that can be transcribed. In
other courts, the public access terminal may only be a name index.
As
the public may not access physical court files directly, name matches must
be brought to the clerk of court staff, who will research them as time
permits and provide the results when they can get to them. In other courts,
searches must be done by hand in old ledger books or paper indices.
In
still others, the automated case management software has been in place for
less than seven years, so searches must be done using a combination of
computer terminal access, old ledger books, and reliance on court services
staff.
Roughly 10 percent of all searches may be done using remote access tools
directly into court systems, but this does not always mean faster or
cheaper.
Even aside from some of the specialized hardware or software required by
some courts, there are connectivity issues and poorly designed systems that
dump page after page of docket activity making it a "Where's Waldo?" search
for the actual identity, charges, and disposition data.
There are still other remote systems that serve only as pointer files and
still require the services of a live agent and court services staff to
complete the research.
Adding another level of complexity to all the above is the fact that each
state has its own, unique criminal code, and each court within each state
may have its own procedures for recording and storing information.
There are about 3,200 counties in the U.S., and about 10,000 different
courts.
Some days it seems that all they have in common is that they are
understaffed, overworked, and uninterested in your problems.
9. Building a Background Check Strategy
With an understanding of the challenges of both on-line and court research,
how does one determine the appropriate scope of or a criminal history
background?
To
answer that question, there is one more concept that must be understood.
That is the concept I call "Managed Risk and Criminal
Histories:"
Hiring employees creates potential risk to other employees, the public, and
to the assets of the employer. Past behavior is one of the best predictors
of future behavior.
Any position being filled must be evaluated to understand
those criminal behaviors that add risk.
Because there is no useful single nationwide repository of criminal history
and since it is unreasonable to conduct a search in every court in the
country, a background program should be designed to cover both a reasonable
period of time and a reasonable geographic scope to manage and mitigate
those risks.
Simply stated: You can't do it all, so what can you do?
I
am going to set down some guidelines based on the experience of clients
around the country.
10. Take these guidelines and discuss them with legal counsel
to see how they may best work within the context of your company and your
employees.
11. A Complete Application.
Every criminal history check should begin with a complete application.
This cannot be emphasized enough.
That application should contain a statement of truth in which the applicant
acknowledges that all information is true and complete,
without omissions, understanding that the employer has the right to choose
not to hire or to terminate if they find out otherwise.
This application should ask the applicant to self-report job-related
convictions.
(There are numerous state and federal laws addressing this section that can
not be address here. Once again, consult with counsel.)
Be
sure that the application has sufficient address history to cover the
required depth of background discussed below.
Also be sure that if you are using a third party background company, you
have a separate release and notification form signed by the applicant that
meets Fair Credit Reporting Act (FCRA) standards.
(Note that some courts require date of birth for an identifier, especially
for common names, and the release is a good place to capture that and other
identifying data).
12. Social Security Check.
The next required tool is the social security number check done through one
of the background check sources.
This check serves three important functions.
First, it may help reveal a legal name change when additional names are
associated with a number.
Second, it serves to raise a red flag if (a) the applicant's name is not
associated with the entered number, (b) the number was issued by an
unaccounted for state, or (c) the date of issue does not seem to match the
applicant's age.
Third, it will reveal up to a seven-year residence address history.
This will be crucial in identifying courts to be searched that would
otherwise be missed should the applicant have selective memory.
13. Criminal History Check.
Now you would use a position risk analysis to decide how wide and how deep
the search needs to be.
The how
wide
is
simply how many years of employment or residence addresses do you need to
cover.
The how
deep
is how many years back do you need to go in the court that you are
searching.
The standard for depth of search in a court is seven years back.
Unless one has specific licensing or regulatory requirements to do
otherwise, this should be your baseline.
It
is the how wide that will vary by position.
A
very simple rule of thumb is to start with the current county of residence
as the most basic and minimum search, and add to that as responsibility and
risk increases.
Taking our rule of thumb, the following are reasonable guidelines for
criminal history checks, along with the number of courts for budgeting.
(When budgeting, remember that about 10 percent of courts charge statutory
access fees ranging from $5 to $26.
Your average cost of fees will vary depending on your applicant pool
demographics, but in a true nationwide distribution, you would be safe
planning $2 per search for access fees. Fees in New York state are
especially high, and Los Angeles courts are $5 per name per search.)
14.-Entry level position,
no cash handling, very little or no customer contact, well
supervised--A check of felonies and those misdemeanors indexed at the court
of
current county of residence, seven years deep. Average 1 court searched.
14 A. -Entry level position,
no cash handling, customer contact, well supervised--A check of
felonies and those misdemeanors indexed at the court(s) covering
three years of residence checked, seven years deep.
Average 1.5 courts searched.
14 B. Entry level position,
cash handling, customer contact, well supervised--A check of
felonies and those misdemeanors indexed at the court(s) covering five
years of residence checked, seven years deep. Average 2 courts searched.
15. Mid-management or key holder position--
A
check of felonies and those misdemeanors indexed at the court(s) covering
seven years of residence checked seven years deep. Average 2.7 courts
searched.
16. Unsupervised customer contact,
such as delivery or repair person--A check of felonies and those
misdemeanors indexed at the court(s) covering seven years of residence
checked seven years deep, plus a sex offender's registry check. Average 2.7
courts searched, plus the state registry.
