Does medical school need to take four years?

It’s an open secret in medicine that fourth year medical students do very little medicine or school. So why do we have a fourth year of medical school (M4)? And do we need one?
​
Why M4 exists
Medical school is hard; there’s a lot to learn. And in the first three years of medical school, this learning is very condensed. But here’s the problem:

​        ​This is a screenshot from my M4 course registration page. My experience may not be totally representative because I’m an MD/PhD and attend a school that (at the time) allowed us to start our clerkships at the beginning of M2. But a careful review of the literature suggests my experience isn’t atypical: while some students complete advanced electives and away rotations, most students have at least several months off. And even when students are doing necessary and productive activities (e.g., doing research or interviewing for residency), its not clear why they should be paying medical schools to do things that cost medical schools nothing (in the case of interviews) or that similarly educated generally people get paid to do (in the case of research).
​        I want to emphasize that I am deeply in favor of medical professionals taking time to do stuff that is not medicine. If I had gone straight through college and medical school, I could’ve finished my MD six years ago, and I have zero regrets about not doing this. My view is that when you are in (and paying for) school, you should be learning. To put this more formally, medical schools should aspire to educational efficiency:

​      The reason medical students have so much time off is that they have to apply for residency in September (i.e., at the beginning of their fourth year). This means that students have to have completed most of their medical school requirements—or at least the ones that residency programs want to assess—before then. But residency doesn’t start until July of the following year, meaning there is a 10 month gap between when students finish most of the training that equips them to become doctors and when they can start practicing as such. In short, medical school isn’t four years because you need four years in order to master the knowledge and skills that will prepare you for residency; medical school is four years for logistical reasons.

Why is this a problem?
First, medical school is expensive: the median medical student graduates with $235,000 of medical school debt; eliminating M4 thus has the potential to decrease students’ debt burden by $59,000. This would be beneficial to graduates, but perhaps more importantly, would save the federal government—which subsidizes a growing share of medical training—a lot of money.
​        Second, the $59,000 figure actually substantially underestimates the cost to physicians of M4, because while students are in school, they aren’t earning money working as physicians. M4 shortens physicians’ careers by a year, thereby reducing their lifetime earnings by around $350,000. Moreover, prolonging medical training defers investments students might make in retirement accounts, buying homes, and so on—investments that would accrue substantial value over their lives. Thus, considering only the cost of medical school itself substantially underestimates the cost of M4 to students, which we can think of as being (conservatively) on the order of $500,000 per student. 21,000 medical students graduate each year, so this comes out to about 10 billion in debt and lost wages per year. 
​        Third, some studies have found that “high debt is likely to drive students towards choosing higher paying specialties.” In short, the more debt someone has, the more likely they are to become a dermatologic laser surgeon rather than a PCP. (I haven’t looked into this in a few years, but my sense is that this relationship is relatively weak; i.e., we shouldn’t expect reducing debt burden to significantly increase interest in lower-paying specialties.) That said, increasing debt burden is unlikely to encourage graduates to pursue lower-paying specialties.
​        Fourth, if you’re worried about doctor shortages (which you may not be), requiring medical students to do an extra year of medical school marginally compounds this problem by reducing the supply of doctors by about 2% of the total.
​        Fifth, long training times lead medical trainees to defer or forgo important life projects. Medical training was designed at a time when most doctors were men and their partners were mostly homemakers. Today, a slight majority of medical school graduates are women, and nearly half have a partner who also works in healthcare. This means medical trainees often face difficult choices with respect to, for instance, starting families: either work 80-hour weeks throughout pregnancy and with an infant, spend a quarter of your income on childcare, and enjoy few parental leave protections or defer having kids until after completing training, and experience much higher rates of infertility and pregnancy complications than the general population. Eliminating M4 wouldn’t make a huge dent in this problem, but it might be one of several interventions that could help.
​        Lastly, many medical students may actually regress during M4–after all, if you’re not using medical knowledge, you are liable to lose some of it. So it’s not just that students are paying for a very expensive year off; medical schools may actually be charging them for the opportunity to graduate as worse doctors. (This article beautifully notes that people who take a more intensive course load during M4 may be at “increased risk of excellent performance” as interns.)

Solutions 
I want to first clarify that I don’t think there are any bad guys, here; in other words, there is no medical school dean looking at the status quo and thinking “this is an inefficient and costly way to train doctors, but it’s good for our institution’s bottom line, so let’s make sure we keep things this way.” Rather: (1) medical school has worked this way for a long time, so it’s widely accepted that medical school needs to be four years, (2) it is in fact good for medical schools’ bottom lines to keep M4s around, (3) medical students themselves are often grateful for the time off (and the fact they they don’t have a gap in their CVs in association with it), and, most importantly, (4) reforming this system would require multiple large institutions (e.g., medical schools, residency programs, and the AAMC) to coordinate with each other, and few of these actors are incentivized to do so, since they’re not the primary ones bearing the brunt of this inefficiency (that would be medical students, taxpayers, and patients). 
​        But with that said, I think reform is possible. And if the goal of the reforms is to improve educational efficiency, we can think of possible reforms as falling into three general categories: improve educational efficiency by increasing the numerator (i.e., providing more education) or decreasing the denominator (i.e., reducing costs and/or decreasing training time). Specifically, medical schools could:

1. Provide more education. I’m not a big fan of the idea of requiring medical students to take more courses: if you think that medical students are adequately trained in the status quo, it doesn’t make much sense to add requirements for the sole purpose of improving educational efficiency. Moreover, I (and a lot of other medical students) think people really need the time off. I also worry that external pressure on medical schools to justify M4 might just lead to them padding their fourth year with unnecessary coursework, which would be bad for literally everyone. But if you could incorporate high-yield training into M4, and plausibly use this to shorten or lighten training in residency, that seems like a good—if hard to implement—idea.
2. Reduce costs by not charging a semester of tuition. If medical schools are essentially going to give students a semester off, they shouldn’t charge them. This seems like a fair and easy thing to do. But medical schools might object by claiming that M4 tuitions subsidize the costs of the other years (which seems unlikely, but is hard to disprove). And if they were forced to waive a semester of tuition, they’d probably gradually increase the tuition across the other seven semesters to offset this loss.
3. Decrease training time by giving students the option of graduating after three years. A number of medical schools have created accelerated pathways that allow students to finish in three years, and a third of medical schools are considering creating pathways like this. Many of these programs have the explicit goals of addressing physician shortages in certain areas (e.g., primary care; rural medicine) and alleviating debt burden. I think it would be good for these kinds of pathways to exist at all medical schools, and to be available for students applying into a wider range of specialties. I also think medical schools should face external pressure (e.g., from the AAMC) to develop such pathways. That said, there are some logistical challenges that limit how many students will want to opt into these programs, because they often require students to choose their specialty earlier, provide less time to do other activities (like research), require students to remain at the same institution for residency, and are more common in certain specialties (like primary care and family medicine). However, some of these challenges are being addressed (e.g., at Duke and Penn State, you can complete a three year MD and continue into orthopedics). And I suspect that if more institutions had these programs, medical schools would learn from each other, residency programs would become more open to them, and accelerated pathways would gradually become accessible to a wider range of students.
4. Decrease training time by allowing students to graduate and start their residencies a semester early (i.e., in January). This suggestion is the most radical (and least likely to be implemented) because residency has started around July 1 for a very long time, and allowing medical students to graduate and start six months early would necessitate a massive amount of coordination across medical schools, residencies, and fellowship programs (which also start in July). But bracketing the tractability of this, shortening medical school to 3.5 years would do the most to address the aforementioned problems without creating new ones.

Conclusion 
M4 is a lot of fun. But I don’t think fun should be mandatory, cost hundreds of thousands of dollars, or unnecessarily limit our supply of doctors. Correspondingly, I think medical students should be able to opt out of M4 if they so choose, and that medical schools should continue (and should face pressure) to develop shorter, more cost-effective training pathways.