Editor’s Note: Brad Landin is president and chief compliance officer of Employment Screening Resources, a background check company based in Novato, California. The opinions expressed in this commentary are solely those of the author.
Brad Landin: Uber driver allegedly shot people in Michigan rampage between providing rides. Uber's vetting of drivers under new scrutiny
He says to work, background checks must use overlapping, complex, local searches. It's not clear how thorough Uber's searches are
The recent mass shooting in Kalamazoo, Michigan, in which the alleged gunman apparently provided rides as an Uber driver between shootings, has placed public attention squarely on the issue of the background checks performed on Uber drivers.
Jason Brian Dalton – who is facing murder, assault and firearms charges – apparently had no criminal record, so a background check would not have helped exclude him. Yet, the case raises questions about personal safety in the so-called “gig economy,” where total strangers are invited into your life through software applications to provide a host of services, ranging from transportation and housecleaning to babysitting and acting as a personal assistant.
Background checks are a complex issue. Contrary to public belief, there is no comprehensive database or source of information that provides a perfect one-stop background check solution. Even the fingerprint database maintained by the Federal Bureau of Investigation is far from perfect.
Although the FBI check uses biometric matching, the shortcomings of FBI criminal searches have been well documented. In 2006 the U.S. attorney general revealed that roughly 50% of the records are incomplete and fail to provide information on the final outcome of an arrest.
Employers using background screening firms can and do access millions of records through unofficial private sector databases as a source of possible information. However, these databases are incomplete and subject to both false negatives (meaning a criminal record is missed) and false positives (meaning the record is about the wrong person or that, by law, the criminal matter should not have been reported). In our opinion, this type of database should not be relied upon as a standalone search.
With some 3,200 counties in America to scour, these private sector databases are by no means complete, accurate and up to date. For example, there are a number of states that provide no public access to statewide criminal records, which usually leaves only a small number of county records that are included in the private sector databases for an entire state.
Within the screening industry, these databases are primarily used as a research tool that can lead to the discovery of information, which must be properly vetted as being reportable under the Fair Credit Reporting Act and corresponding state laws. The standard screening industry recommendation is to check primary source courthouse records where a person has lived, worked, or studied for at least the past seven years.
A thorough search does not begin and end with a background check. Employers should utilize a series of overlapping tools to exercise proper due diligence. Often, procedures followed before the background check may be the most important: a thorough review of the application, verification of past employment, accounting for gaps in employment and conducting an in-person or video interview before a making a final offer.
In the gig economy there are so many moving parts that it is hard to know just how much due diligence a company is using. For example, if a service engages someone through an online application without ever meeting him or her, there is the question of whether the person being researched is actually the person who is applying for the job.
In a standard employment situation, the employer – at some point – meets the person and engages in some sort of identity verification. That may not always be the case in the gig economy, where people may sign up through a software application, sight unseen to provide ad hoc services.
Uber has published a detailed description of its screening program here.
To a background screening professional, there are some unanswered questions about this program. For example, Uber claims it searches numerous databases and then does a courthouse confirmation to ensure accuracy.
However, the policy does not explain how Uber deals with those counties where databases are unavailable or do not provide actionable information. A solid due diligence approach generally includes searching counties relevant to the applicant, using primary source information. Using only a database is not the functional equivalent of going to each county courthouse related to the applicant, which leaves the question of whether all applicable counties are being properly searched.
Uber states that it has a seven-year search policy. Even assuming seven years is the proper time limit, the issue of how the seven years is calculated can be complex. A driver may have been convicted and gone to prison 12 years ago and been released from custody three years ago.
If the scope is limited to just the past seven years, an older criminal matter can be overlooked, even though it may be reportable and relevant. In California for example, convictions can be reported from the date of conviction, release or parole. However, San Francisco counts exclusively from the date of the offense.
But, again, safety goes beyond just the background check. Third-party gig economy application providers such as Uber, Lyft and others like them can build in safety protocols that can be very effective for their services.
Background checks are complex, and consumers are entitled to know exactly what measure of safety they are getting.
That said, it is critical to understand that no matter how much due diligence is performed, no system is perfect. In the case of the Kalamazoo driver, a criminal check would have apparently come up clear.