Unemployment That Doesn't Suck: Part 2 - Conduent
What does the biggest company in the space actually get out of administering EPC benefits?
If you haven’t read Part 1, Part 2 won’t make sense. Catch up here :)
A Brief Corporate Overview
As I mentioned last time, the big private player here is Conduent. Which is really a new-separate division of Xerox, which for a while wanted to make more money from non-printer things before the Street told them that wasn’t working. Go figure. In 2022, both Conduent and Xerox posted a net loss of about the same percent of revenue. The idea that 2 smaller, more focused companies could succeed where a conglomerate couldn’t doesn’t seem to have played out.
The split occurred in 2017, and even at that time the company counted all 50 states and 500+ other governments among its customers and claimed to be the largest prepaid card processor in the world. Their share, by all accounts, has certainly not decreased in the intervening 6 years.
It’s tough to tell, though, because despite being the worldwide leader, their most recent 10-k never mentions the service. That’s especially strange considering that their governmental services segment is it’s 2nd-largest by revenue and largest by EBITDA generation.
How Conduent Makes Money
At the end of the day, Conduent is a software services company and makes money the same way that whole sector does: big contracts for the rights to use their software. The golden metric here is Annual Recurring Revenue (ARR), or the amount of revenue that’s under contract to repeat each year. For Conduent, that’s over $100M.
The segment we care about here is what they call “Government Service Solutions”, and per their most recent 10-k here’s how they define their business:
With $106 billion disbursed annually, we are a leader in government payment disbursements for federally sponsored programs including Supplemental Nutrition Assistance Program (SNAP), formerly known as food stamps, and Women, Infant and Children (WIC) as well as government-initiated cash disbursements such as child support and Unemployment Insurance (UI). We deliver electronic payments for government services in 36 states, 23 Electronic Benefit Transfer (EBT) programs, 13 EBT for WIC programs and 6 Electronic Child Care programs. In our SNAP payments solution, we generate revenue based on the number of cases or number of card holders. Within our UI payment solution, we generate revenue based on interchange fees and spending on cards as a percentage of transactions.
They also have a separate section specifically defining their support of child support programs, so it’s fair to read the above as relating to support for all non-child support programs.
Their revenue generation depends on the type of program:
SNAP fees paid to Conduent are essentially “per user” fees, which likely sounds familiar if you’ve ever bought any kind of enterprise software. The more people enrolled in SNAP, the more money Conduent makes. They don’t really care about how people use the program, just that they exist and are active.
UI fees paid to Conduent are explicitly usage-based, though - so, they don’t care how many users are out there, they care how they’re actually using the product.
In general, for something like this, usage-based payment is likely better for everyone. It keeps incentives aligned, where the provider is only paid if the solution is actually working. But, in this case, it’s worth diving a bit deeper…
…Interchange fees and spending on cards as a percentage of transactions…
If you’d asked me before I started researching to define the difference between these 2 things, I wouldn’t have been able to. And, frankly, even after researching I’m not sure that I’m right. But, here’s my best stab at working definitions for these terms:
“Interchange Fees” are transaction fees that a merchant's bank pays to a card-issuing bank whenever a customer uses a credit or debit card to make a purchase that exist to cover the risk of fraud and handling costs for actually sending the payment.
When I talk about “payment fees” in other posts, this is what I’m talking about.
“Spending on cards as a percentage of transactions” just means that a percentage of each transaction value gets paid back to the provider.
The real question I have is why these are separated out. Outside of the nominal flat interchange fee (usually 30 cents), they’re always assessed as a percentage of the transaction. And my guess, though I can’t confirm, is that Conduent separates them out because their interchange fees are likely always “market rate” while the additional percentage is what’s customized to any given customer/solution.
But, at the end of the day, Conduent makes money on an EPC they administer when someone actually uses that EPC.
A Problematic Incentive Structure
Only making money when a specific format of a program is used means that you’re incentivized to maximize that specific format. That’s why the default UI format in Arizona is the card and not direct deposit, and that’s a problem. This kind of structure takes away choice in favor of a suboptimal default value that’s better for the provider but not for you.
If, instead, payment to Conduent was based on number of unemployment claims filed regardless of how they were fulfilled, then Conduent wouldn’t care who got direct deposit and who used a card. There’d be a clear choice to make on the application and no hoops to jump through.
That would be a harder contract to negotiate, because there’d be so many more assumptions; how many people choose which format, how many transactions are card users making, etc. Much more complicated, and more risky, than just saying “5% of every transaction and that’s it.” And there’d be additional costs inherent in managing something more detailed like that.
In short, neither existing structure really works for all parties involved. But, at least now I know for sure why a debit card I didn’t know existed just showed up in the mail for me one day.
What’s even more problematic is that even though I know in general what fees my state is paying, I have no idea of the specifics. Tax dollars being spent without any real traceability isn’t a recipe for success.
Next week, I’ll come back and discuss a different structure and how we might be able to increase transparency and fairness in the administration of these critical programs.