Having an IVF Baby vs Buying a Toyota
…or Getting into Harvard Med School
Harvard thought it was special. It was the only med school that charged a fee to send you an application. I think it was $20. Back then I could withdraw $5 from the George machine at the Girard Bank at 36th and Walnut on the edge of campus and it would last a week. So I figured to hell with Harvard.
I stayed at Penn.
No one wants to buy an IVF cycle. An investor asked me once, not in an obnoxious way, what the big deal was about IVF access. IVF costs about the same as an entry-level Toyota, no?
He was right, but the difference is that when you bring $21,700 into a Toyota dealership, you drive out with a Corolla. The same $21,700 at an IVF clinic may pay for a 50% chance of having a baby, so the effective cost of buying what you want is closer to $40,000.
But it’s worse than that. The unfortunate 50% who fail to conceive have *nothing* to show for the money spent, other than having gone through three weeks of living injection to injection and other pains in the butt, literal and figurative.
I spend a lot of time modeling demand for IVF, and I get criticized for what seem like unrealistically large addressable market sizes. If my model, for example, predicts United States demand for IVF sufficient to produce 700,000 or 800,000 or a million babies per year, when IVF currently accounts for less than 100,000, then my model must be structurally flawed, no? Take infertility alone. Simply extrapolating from the prevalence to demand for the treatment — must overestimate the percentage of affected people who actually want to do IVF, right?
Well, yes and no. The 100,000 successful IVF patients had both the ability to access IVF and made the choice to buy the lottery ticket that gave them a shot at a baby, and was able to stomach the possibility that at the end of the process, they would have set fire to months of salary or the proceeds from a second mortgage on their house — and come away with nothing.
Modeling IVF consumer behavior needs to account for not only the three metrics of dollars per baby, time to baby, and life disruption until baby that we talked about a few months ago:
IVF demand = 1 / ($ / baby + t / baby + LD / baby)
It also needs to be taken into account that each patient has a different attitude about a very real chance of spending tens of thousands of dollars to come away with nothing other than the sunk cost burden of feeling like you have to spend tens of thousands of dollars more to justify having spent the money the first time.
IVF demand = 1 / (($ / baby + t / baby + LD / baby) + (can’t tolerate wasting $ if it never works))
As the price of successful IVF decreases, as we engineer it to scale (everyone catches up to the results of the best clinics — which they inevitably do as technology-based industries scale), both parts of the right side of the equation make IVF more attractive. $ / baby and time/baby are obvious. Hopefully, life disruption decreases with a process-optimized cycle too.
And the fear of spiraling down a path of setting tens of thousands of dollars on fire one month after another and feeling you have to keep doing it until it works just, not only to have the baby you desperately want but also to justify the sunk cost of the money spent already — that cost, on a probability basis, is lower too.
Translated into economist-speak, the elasticity of demand for IVF changes as the chance of conceiving each time is not linear — demand increases at an increasing rate as the cycle outcomes improve.
This means that looking at past IVF utilization probably underestimates the demand for future use when the probability of outcome is higher, out-of-pocket costs are lower, care is more easily accessible, and risk of financial insecurity or ruin is alleviated by simpler underwriting and risk management from a more predictable process.
There are reasons why we spend less on IVF worldwide than we spend on potato chips (about 22 billion dollars vs. 27 billion — look at the financial statements of the Frito Lay division of PepsiCo.) How badly people want to have children is not one of them.
