Gig workers don’t have good health benefits. That’s not news. But the usual suspects-cost, regulations, lack of employer interest-aren’t the whole story. The real culprit is something most people never see: the software that runs the entire benefits system.
Think about it. Every major platform-Workday, ADP, Benefitfocus-was built for a world where one person works one job for one employer. That world is fading fast. Today’s workforce is fluid, multi-platform, and unpredictable. But the administration systems haven’t evolved. They’re like trying to run a smartphone app on a 1980s mainframe. Here’s where the breakdowns happen, and what we actually need to fix.
The Data Model That Doesn’t Fit
Every benefits system is designed around a single core entity: the Employee. Under that sits a table for Spouse and Dependents. It’s a neat one-to-many relationship: one employer, one employee, a few family members. That works fine for a traditional 9-to-5 job.
Now picture a gig worker who drives for Uber, delivers for DoorDash, and does a few hours on TaskRabbit each week. She also has a part-time W-2 position at a retail store. The system doesn’t know what to do with her. There’s no field labeled "Primary Subscriber with Multiple Funding Sources." The software forces a binary choice: pick one employer as primary, even though her income comes from four different places.
The result? Confusion, errors, and a lot of manual workarounds. The system could handle the situation; it was simply never designed to.
- The fix: Switch to a worker-centric data model. Instead of anchoring everything to an Employer ID, attach benefits records to a portable digital identity-a kind of benefits wallet.
- What that means: The system would accept contributions from multiple platforms in real time, like 30% from Uber, 20% from DoorDash, 50% from the worker. No single employer owns the coverage.
The Coverage Continuity Trap
Group health plans run on a simple binary: Active or Terminated. The moment a gig worker finishes a ride or a delivery, their "employment event" ends. In a traditional system, that triggers a termination notice. Then, when they pick up another gig the next day, it’s a new hire event.
Gig workers exist in a permanent state of eligibility whiplash. Every week the system sees a day-one termination followed by a day-two new hire. That means constant COBRA letters, new ID cards, and re-underwriting. It’s a nightmare for administrators and a huge coverage gap for workers who actually log 40 hours a week-just across 50 different micro-tasks.
- Problem: The system treats horizontal liquidity like vertical job hopping.
- Solution: Implement hour-banked or task-banked eligibility. Track aggregate hours or earnings across all platforms.
- How it works: When a worker crosses a threshold-say, 120 hours in a month-the system auto-enrolls them for the next month, retroactively if needed.
This requires real-time API integration between gig platforms and carriers. Right now, almost no major benefits system supports that natively.
The Subsidy Calculus Blindspot
Traditional benefits systems are excellent at handling pre-tax payroll deductions and fixed employer subsidies-the company pays 80% of the premium, deducted from a W-2 paycheck. Simple.
But some arrangements already use a different model: a variable per-transaction contribution. The Black Car Fund has run one since 1999, funding workers’ compensation for more than 80,000 New York for-hire drivers from a small percentage added to every passenger fare. That works in practice, but a conventional benefits administration system has no native module for a variable-contribution fund. A single $500 monthly premium could require reconciling contributions from thousands of individual rides.
Because the back end can’t handle it, platforms often just pay the money out as ordinary income. The worker loses the tax advantage, and the money never becomes coverage. Everybody loses.
- The fix: Create a premium escrow account with algorithmic settlement. The gig platform sends a daily streaming file of contributions. The admin system holds the money until the worker’s premium is due, then pays the carrier.
- The catch: This requires robust KYC (Know Your Customer) and AML (Anti-Money Laundering) logic built into a benefits system-capabilities that currently live in FinTech, not BenefitsTech.
The Middleware That’s Missing
The unique insight here is that plan design alone doesn’t solve this. The binding constraint is data plumbing: a benefits switchboard that sits between gig platforms (as funding sources) and carriers (as risk pools).
This middleware would do three things that mainstream HRIS and benefits administration systems still don’t do:
- Identity de-duplication - recognize that "Jane Doe" is the same person on Uber, DoorDash, and Instacart.
- Aggregation and pro-ration - calculate her total gig income and automatically split the premium across platforms based on her earnings percentage.
- Real-time status updates - tell the carrier she’s eligible the moment she hits the threshold, not via a weekly 834 file.
The Switchboard Has Already Been Prototyped
The pieces this post describes aren’t hypothetical. The Black Car Fund has run a working version of the switchboard since 1999. A small passenger surcharge on every New York for-hire trip flows into a single fund that provides workers’ compensation to more than 80,000 drivers, including Uber and Lyft drivers, plus a suite of health benefits that drivers enroll in separately. Per-transaction funding and centralized aggregation have been in production for a quarter century.
The catch is scope. The Black Car Fund works because a state statute created it and requires every base to collect the surcharge. Alia, a portable benefits platform from NDWA Labs, showed the model could work for scattered domestic workers starting in 2018, letting clients contribute to paid time off and insurance; the original platform wound down in 2022. On the policy side, the Portable Benefits for Independent Workers Pilot Program Act was reintroduced in Congress in 2026 with $20 million for state and nonprofit pilots, and a separate bill would clarify that offering portable benefits does not, by itself, make a worker an employee.
Technology is only part of the story. The harder problems are governance and incentives: who runs the fund, who is required to contribute, and who gets the data. A benefits switchboard is infrastructure, and infrastructure like this tends to get built when a law or a dominant platform forces everyone onto the same rails.
The Bottom Line
The conversation about gig worker benefits is stuck on politics and cost. But the real fight is in the data architecture. Until we build systems that treat the individual as the center of the policy-not the employer-we’ll keep falling short. The deeper issue is a system that can’t compute the reality of modern work.
The ghost in the machine is W-2 legacy code. It’s time for an upgrade.