Restrictions often exclude workers from the gig economy while increasing property crime. Here’s what platforms, communities, and regulators should do instead.
Authored by CHEN Yanzhen
When she was in her 20s, Ingrid Archie was convicted of drug possession. Years later, despite having worked for her company for three years, she was laid off when a policy change barred anyone with a criminal record. Suddenly jobless and supporting a young child, she turned to theft, stealing clothes for her baby—and was promptly rearrested.
Ingrid’s story is far from uncommon. In America, one in three adults has a criminal record, and the vast majority of employers require background checks. As a result, millions of people find themselves stuck in a cycle of poverty and recidivism.
Fortunately, Ingrid’s story had a happy ending: While her criminal record limited her access to traditional employment after her release from prison, she was able to get a job driving for Uber. This helped her launch a new life, enabling her to break the cycle and achieve lasting stability for herself and her family.
Because of their low barriers to entry and flexible work models, ride-share platforms like Uber and Lyft often end up being a lifeline for people like Ingrid, who might otherwise struggle to get a job due to their criminal pasts. However, new laws mandating background checks even for gig workers threaten to erode this vital benefit of the gig economy without necessarily even benefiting the communities these laws are meant to protect.
Background Check Laws Often End Up Backfiring
To be sure, background check laws, or BCLs, serve an important purpose. After all, the same accessibility that makes rideshare work attractive to people with criminal records increases the likelihood that passengers will be exposed to crime. However, our recent research suggests that many BCLs end up backfiring, unnecessarily keeping low-risk people from accessing one of the few avenues for legal work available to them.
With no other options, these people end up falling back into criminal activity, ultimately increasing overall crime rates.
Specifically, to explore the true impact of BCLs, we analyzed thousands of criminal reports and docket cases (formal court filings) involving rideshare drivers. We excluded cases related to vehicle accidents and labor disputes, but captured allegations ranging from theft and robbery to physical and sexual assault, unlawful restraint, and other criminal incidents directly tied to rideshare trips. We then leveraged the variation in BCLs passed in different states at different times to determine the impact of these laws on both rideshare- related incidents and crime rates more broadly.
Through this comprehensive analysis, we found that after BCLs were adopted, rideshare-related docket cases declined significantly. In other words, as expected, excluding higher-risk drivers is effective in preventing rideshare-related crimes. But at the same time, we found that property crimes in communities with new BCLs increased by approximately 6%, suggesting that displaced workers were pushed toward other offenses.
Digging deeper, we found that not all background check laws had the same effect. In a follow-up analysis, we explored the impact of more and less strict BCLs, and we found that stricter laws— that is, regulations that excluded drivers for non- violent, non-recent offences such as identity theft, fraud, or evading police (even if these convictions were more than five years old)—led to worse outcomes than laws that only targeted recent, violent, or sexual crimes: These stricter BCLs did not reduce rideshare crimes any more than the less-strict laws did, but they significantly increased property crime rates.
This analysis demonstrates that in addition to the direct social and financial harm they cause to people with criminal records, strict BCLs often increase overall crime rates without even improving outcomes for the passengers these laws are designed to protect. So, what will it take to effectively protect customers’ safety without excluding good workers and unintentionally increasing crime rates?
Three Theories of Criminal Activity
Designing better interventions starts with understanding why people commit crime. Criminology theory points to a range of drivers: First, Routine Activity Theory suggests that crime occurs when a potential offender is motivated, encounters a suitable target of crime, and that target is insufficiently protected. In other words, crime happens when a person’s routine activities enable and incentivize it.
Next, Labeling Theory argues that stigmatizing people as “criminals” can reinforce cycles of exclusion and recidivism. Whether that’s a government regulation that forces people to disclose their criminal past or a voluntary organizational policy or social norm, the stigma of being labeled as a criminal can make repeat offenses all the more likely.
And finally, Rational Choice Theory posits that people weigh the benefits and costs of crime versus legal work, and when legal options come up short (or feel altogether inaccessible), crime becomes the rational choice. Taken together, these theories shed light on why indiscriminate background checks may inadvertently fuel more crime than they prevent—and what other approaches might be more effective.
Effective Deterrence Takes More Than Background Checks
Specifically, these theories and our own research inform several strategies that platforms, communities, and governments can use to deter crime and protect communities (without some of the unintended consequences of BCLs).
1. Platforms Can Leverage Technology to Deter Crime
For platforms, there are a range of tools that can help ensure passenger safety without excessively excluding drivers. For example, innovations such as license plate verification (in-app notifications reminding riders to check their car’s license plate before entry), real-time ride tracking (the ability to share trip status with trusted contacts), emergency button integration (instant 911 access with location data), and unusual stop detection (AI alerts for suspicious detours) all help boost rider safety without restricting drivers. Indeed, our study showed that when Uber rolled out these features, crime rates fell by up to 30%—a much larger effect than BCLs alone.
2. Platforms Can Avoid Blanket Bans
Rather than screening out anyone with any kind of a criminal record, platforms can screen variably depending on the job requirements. For jobs that are higher risk, it may make sense for platforms to screen more aggressively. But for many food delivery and service apps, less restrictive screening is sufficient to protect customers while ensuring that more workers have access to these important alternative income streams.
3. Communities Can Support Workforce Reentry
Local communities also have an important role to play in supporting both safety and job access for people with criminal records. In particular, nonprofits and agencies can invest in programs to help returning citizens secure housing, job training, and counseling. For example, the Fortune Society in New York and the Safer Foundation in Chicago offer a range of resources and support to help people reenter into the workforce.
4. Governments Can Provide Social Safety Nets
Finally, national, state, and local governments can all take steps to ensure that their citizens have access to reliable social safety nets. Financial cushions such as unemployment benefits or other forms of welfare reduce the immediate incentive to return to crime when BCLs limit access to legal work. As such, rather than just passing restrictive legislation that can end up pushing people back into illegal activity to survive, providing support through programs like these can help people reintegrate productively into society.
Building a Safer, More Inclusive Gig Economy for Everyone
Part of why BCLs can be particularly harmful is that they tend to disproportionately affect marginalized communities. Due to systemic inequalities such as racial and socioeconomic disparities in sentencing and reentry, laws that target people with criminal records often also end up worsening a wide range of outcomes related to inclusion and equity. As such, understanding these mechanisms and dynamics is vital for policymakers, platforms, and communities to design interventions that will maximize safety and wellbeing for everyone.
Of course, BCLs will still have a part to play, as effective regulation is an important way to protect customers from potentially risky drivers. But our research demonstrates that these restrictions can sometimes backfire, especially when they’re excessively strict. As such, complementary strategies such as in-app safety features to deter and monitor crime and community or government programs to support reintegration can help ensure customers are protected without excessively excluding workers and unintentionally pushing them back toward crime.
This article draws on the research paper “Exclusion for Public Safety or Inclusion for Gig Employment: Managing the Tension with A Trilogy of Guardians,” authored by Arun RAI, Chen Yanzhen, and LIN Yatang.
Chen Yanzhen is an associate professor of information systems at HKUST, specializing in emerging technology, causal inference, and statistical learning.