The main problem with the first version of Even boils down to a logical fallacy.
“Denying the antecedent,” A.K.A. “fallacy of the inverse,” takes the following form:
If P, then Q.
Therefore, not Q.
(“Denying the antecedent,” 2016, para. 1)
Perhaps you can see for yourself what’s not quite right there. But turning it into a little story makes it even more obvious; take this argument about our coffee consumption at Even:
If it’s morning, then we’re drinking coffee.
It’s not morning.
Therefore, we’re not drinking coffee.
In fact, we drink coffee all day. The argument is invalid. Here’s another:
If the windows are open, there are flies in the office.
The windows aren’t open.
Therefore, there aren’t flies in the office.
Actually, flies remain in our office long after the windows have been closed. They cease to be only once we swat them with our electric fly-swatter. This argument is also invalid.
Back to Even v1. We knew from existing research that income volatility creates periods of income scarcity (Morduch & Schneider, 2013), and that scarcity makes it harder to do the type of planning that leads to financial health (Mullainathan & Shafir, 2013). So, we decided the best way to increase the financial health of people with volatile incomes would be to eliminate their income’s volatility.
We did this by creating the “Even Salary” — a set amount of money, roughly equivalent to the average amount a member earns per paycheck, that Even would guarantee they’d have in their bank account every payday. Once people knew they could rely on the same amount of income each pay period, we theorized, it would be significantly easier for them to plan, and they would do so much more successfully. This argument can be rephrased as:
If your income is volatile, you can’t plan.
Your income isn’t volatile [because now you have Even].
Therefore, you can plan.
The premises of that argument are true, but it contains a fallacy of the inverse: just because volatile income makes it hard to plan doesn’t mean stable income makes it easy to plan. Other conditions might be necessary for ease of planning. Actually, we thought they almost certainly were.
That’s why we saw the Even Salary as just one of the building blocks of the eventual Even product — the H to its H2O. The H of smooth income might be powerful on its own, but it would have its biggest impact when paired with an O of steady spending & saving.
We really believed in H2O and were tempted to just put our heads down and build it. But we also had doubts about whether we could even successfully create H, and we wanted to get confirmation that we could before jumping into building the full molecule. It would be quicker and easier to build H compared to H2O. And it would allow us to start learning from people earlier.
So that’s what we did. We created a basic first product that guaranteed each Even member a steady “salary” by depositing extra money to their account when they got a below-average check and withdrawing extra money when they got an above-average one. H was born.
Again, our standard for the success of H was whether it helped Even members plan out their finances. One key signal of their ability to plan would be how easy it was for them to let Even withdraw the extra when they got high checks.
We weren’t sure what to expect, but we figured that once members started feeling like they were living on their new Even salary number, they would stop counting on any “extra” they earned above that in their high checks, and the withdrawals from those would be a relatively painless, automatic process. Well, we were wrong.
Starting out, we heard a lot from members about how great it was to get extra money to boost low checks — how this allowed them to pay bills they otherwise would have been late on, to cover necessities like gas and groceries which they would have otherwise had to do without — essentially, to stick to their plans.
But when it came to us withdrawing extra money when people’s checks were high, we often got reactions like this:
In many cases, people still ended up needing the “extra” money in their high checks, despite having earlier received extra money to supplement their low checks. Knowing they had a steady amount to live on clearly hadn’t made it easy for them to live on that amount.
Most of these members had started out as enthusiastic about planning as we were. Shantel,* an hourly employee of a grocery store chain in Atlanta, Georgia, was one of the first people to use Even. When I interviewed her shortly after she signed up, she was speaking our language. “Knowing what your paychecks are before you get them necessarily gives you the opportunity to plan out what needs to be paid this week,” she told me. “What money can I save, what money can I hold onto to go shopping or whatever the case is.” Based on her new Even salary amount, she reported, she’d gone ahead and made a budget for the next ninety days.
But despite all that planning, Shantel wasn’t always able to let Even withdrawals from high paychecks go through:
Sometimes it was because she needed the extra to cover her fixed expenses:
Sometimes it was because, following a pre-Even habit, she’d worked overtime with the explicit goal of using the extra pay for a large upcoming expense:
Shantel was keen to pay us back and to save, and she did both proactively:
But the schedule on which she could transfer money out of her account to Even was highly dependent on the still-unpredictable nature of her expenses. She was still living with the stress of doing constant mental math about what she could pay when — with that “what” now including Even. Even had given her liquidity, which she valued highly — but it hadn’t truly given her stability. Plain old H was no H2O.
Though we’d assumed from the start it was a fallacy to say stable income guarantees easy planning, the experience of building H taught us just how fallacious it was. It showed us that as useful as hydrogen can be, it just won’t do when what people really need is water. It made us eager to get back into our chemistry lab and create H2O.
* Shantel gave me permission to use her story in this post.
Illustrations by Marvin Astorga
Denying the Antecedent. (n.d.). In Wikipedia. Retrieved March 30, 2016, from https://en.wikipedia.org/wiki/Denying_the_antecedent
Morduch, J., & Schneider, R. (2013). Spikes and dips: How income uncertainty affects households. U.S. Financial Diaries, 1. Retrieved from http://www.usfinancialdiaries.org/issue1-spikes
Mullainathan, S., & Sendhil, E. (2013). Scarcity: Why having too little means so much. New York: Henry Holt and Co.
Get updates around new research and findings in your email.