What's a point on the MCAT worth?

If you had the money (or the borrowing power) what would or should you pay for an additional point on the MCAT? That depends on what it's worth to be a doctor, and how the incremental point affects your chances.

There's no way to estimate the intangible value of getting into medical school, but it is possible to estimate the financial value, and I've given it a shot. The answer, based on what I believe to be reasonable assumptions (which I detail below):

For a candidate whose MCAT score is in the marginal range, a single additional point is worth somewhat over $10,000.

It turns out the calculations are powerfully sensitive to some of the assumptions, so I will lay out them out in detail for those who doubt, or who want to play around with them. I'm also making the calculations available to everyone who wants to look at them: just click here for the google spreadsheet.

The toughest part of the calculation is estimating how a single point affects one's chances, but I solved that as follows: First, consider the lowest score you could reasonably imagine getting, and estimate your probability of getting in. That number will be near zero for many MCAT takers. Next, estimate the maximum score you can imagine getting, and again estimate the probability of acceptance. The calculation works well for strong candidates, i.e., those for whom a good enough MCAT score will get them interviews, and who will have a good chance at acceptance if they get interviews. Now, because it's probably not possible to estimate where you will fall in that range, use the average: the average incremental probability of acceptance is the difference between the probabilities, divided by the score range.

To illustrate, an example: Suppose the lowest score you can imagine getting is a 22, and the highest you can imagine getting is a 32. (conservative assumptions; a tighter range increases the value) Suppose that with a 22 there's some chance you'll still get in, call it 10%. With a 32 there's a good chance you'll get in somewhere, maybe 70%. There's a 60% difference between the chances, and a 10 point score range, so on average, an additional point increases your chances by 6%. (Some points will be worth more, some less, but by hypothesis you don't know where in the range you'll fall.)

Using my first set of assumptions I calculated a net present value per point of about $11,900, though this level of precision is silly. Really, the importance of the calculation is not to show what the number is, but to show its order of magnitude.

The assumptions I used are detailed in my next post.

 

My first calculations produced a value per point of $11,919. Here are the assumptions I used:

Scores and acceptance
Score range (highest to lowest conceivable scores): 10
Chance of acceptance with lowest conceivable score: 10%
Chance of acceptance with lowest conceivable score: 70%

Social
Med school is four years long, plus residency of four years
Candidate is 22 years old, and will work until age 65.
Alternative to medical school is starting work immediately (not graduate school)

Income and cost
Gross income as MD = $147,000 (the national average)
Gross income as resident = $45,000
Gross income as non-doctor starts at $40,000, rises to $80,000 over 15 years
Medical school cost/year = $30,000 (note: does not include cost of living, because you pay that regardless)

Other financial variables
Discount rate (roughly the same as the interest rate over and above inflation) = 5%
Total tax rate on income = 45% (realistic, when all taxes are included)

All calculations are in today's dollars.

The results are quite sensitive to assumptions about income -- if you're going to go into a high-paying specialty, the value goes way up; if you have great job prospects outside medicine (without any more school), they go way down. They are not very sensitive to tax rate, discount rate, or medical school cost.

Feel free to grab a copy of the spreadsheet and play with the assumptions yourself.

 

The problem with that approach is exactly what you mention: higher MCAT scores are definitely correlated with greater strength in other areas. That's the kind of approach I would have taken when I was at Booz-Allen, where we were paid a lot of money to use whatever assumptions it took to prove the point our client wanted us to prove.

In any case, I wasn't looking for precision, but a relatively unassailable, very conservative order of magnitude calculation.