Who Would Find This Article Most Helpful
- Those who find going to bed at a reasonable hour a major bottleneck to getting enough sleep and maintaining a healthy sleep schedule
- Those interested in thinking more deeply about their mindset with respect to sleep and productivity
TL;DR
Things I’ve Tried Shortened
- Paying someone else $0.01 for every minute later I go to bed than my bedtime. (Perhaps next time I could use SPAR).
- FocusMate while doing bedtime routine.
- Posters reminding myself to sleep.
- App/Website blockers.
- Reading Why We Sleep.
- Making rough calculations on productivity loss.
Things I am Currently Trying
- App/Website blockers.
- Listing out the ways I fail to go to sleep earlier and strategizing ways to combat those failure modes.
Mindset Shifts
- Framing staying up late to meet deadlines as a high-interest loan.
- Recognizing planning fallacy with variability in how we feel the next day while being sleep deprived.
- Realizing that giving up sleep to be successful backfires.
Recommended Exercises
- Reading Why We Sleep.
- Making rough calculations for loss of productivity on sleep-deprived days.
- Enumerating and combatting failure modes that lead to sleep procrastination.
Helpful Resources
- Read this article to find them!
My Personal Sleep Procrastination
How I Procrastinate on Sleep
I procrastinate on going to sleep in a wide variety of seemingly unrelated ways:
- Downloading software
- Having long phone calls
- Over-engineering spreadsheets
- Listening to good music
- Cleaning up my room
- Having messaging exchanges
- Reading cool articles or a good book
- Finishing up homework
- Making posters
- Ironically, working on this article
- And many more!
Things I’ve Tried
- Paying someone else $0.01 for every minute later I go to bed than my bedtime. It worked in the sense that my sleep schedule became better, but I ended up losing enough money that it was painful enough for me to discontinue this. Having a sleep accountability buddy system could work better (two people paying each other, so that no one loses too much).
- FocusMate nightly routine sessions. I ended up cleaning my email instead of brushing my teeth, showering, and actually going to sleep.
- App & Website Blockers. I kept quitting the blockers. I am currently experimenting with a script (for Freedom on Mac) that reopens the blocker every five minutes, which seems quite effective. I do sometimes go on my Phone/iPad instead to avoid the blockers.
- Listening to the audiobook Why We Sleep, but I still find going to bed at a reasonable hour hard.
- Posters reminding myself to sleep. Didn’t work but probably was helpful towards shifting my mindset a little bit.
Given that a lot of these interventions didn’t work, I hypothesize that shifting my mindset about sleep—even though I know it is quite important—is probably necessary to ensure that my system of tools actually works.
Why Sleep is Hard & Making the Mindset Shift
1. Staying Up May (Sometimes) Be the Optimal Choice
There are times when we do not regret staying up late. In fact, in the short-run, staying up late may well be the optimal decision. Here are some examples:
- Preparing a presentation that you need to present the next morning. It would be extremely unprofessional to have many visibly unfinished slides.
- Finishing up an assignment that is worth many points and has a hard due date, so that maybe you get a 90% on the assignment rather than a 50%.
- Studying for an exam (sometimes). For instance, in my experience, it could be better to learn 5 new Spanish vocabulary words and be slightly more sleep-deprived the next day than having learned 0.
The general theme is that there is a hard deadline that one has to meet and that it would be much better in the short run to meet it than to get the extra dose of sleep.[1]
The Problem
We might extrapolate from the experiences where staying up later would be the optimal choice and assume that we should stay up late to be more productive and get more done. However, forgoing sleep could be thought of as a high-interest loan.[2] You’re likely to be less effective and feel awful the next day. In addition, it takes roughly 4 days to recover 1 hour of sleep debt in terms of performing at the optional level.
2. Variability in How We Feel the Next Day
Maybe sometimes you stayed up late and did something productive, and when you woke up the next day you didn’t feel that bad.
The Problem
With planning fallacy, you think that you’d end up ok tomorrow even if you sleep late tonight when that is not always the case.
3. Stories of Successful People Who Sleep Very Little
Elon Musk. Steve Jobs. Nikola Tesla. Thomas Edison. (Maybe) your friends at elite universities.
The Problem
Our careers span decades. Maybe being sleep deprived for a few years can work out, but this is unsustainable in the long run. Steve Jobs died young. Nikola Tesla wrote love letters to his pigeon. Elon Musk’s tweets suggest that he may not be thinking clearly. Meanwhile, Jeff Bezos gets a full 8 hours.
In addition, correlation does not imply causation, and we cannot extrapolate from the habits of successful people due to survivorship bias; there are many people who have slept little but were not successful. Numerous studies suggest that more sleep increases productivity and sleep deprivation is a major risk factor in terms of developing burnout.
4. Wanting To Accomplish Something
If you don’t do X tonight, you don’t think you’ll ever do X. Or perhaps you feel like you haven’t been productive for the rest of the day and have an urge to stay up late to catch up on some work.
The Problem
Is doing X more important than getting enough sleep? Does doing X even matter at all? If it was actually that important would I really never do it? Is it possible that you were unproductive during the day because you were sleep deprived?
Recall that numerous studies suggest that sleep increases productivity and decreases the risk of burnout.
5. Insomnia
If you probably won’t fall asleep anyway, might as well go to bed later.
The Problem
I’ve definitely experienced a lot of insomnia. I have plenty of recollections of being awake at 5am when I went to bed at a much more reasonable time. There are many great articles on how to fall asleep more easily. Personally, I found that wearing these blue / green blocking glasses[3] a couple of hours before sleeping (still experimenting with exact timing) and taking magnesium glycinate supplements[4] right before bed is a very low effort way to mostly get rid of my insomnia, though this may or may not work for you.[5]
Exercises To Try
1. Reading Why We Sleep[6]
Realizing that sleep is the foundation for optimal performance and good health (ability to eat healthy and desire to exercise are both impacted) has led me to write this post and want to prioritize sleep. An easy way to do this is listening to an audiobook while going on walks or doing chores like cooking.
I realize that there are many critiques on this book, but if you know that your sleep deprivation is actively harming you, which it was in my case, reading it is a net positive. Sometimes seeking out something helpful or useful in the short run may mean sacrificing a little bit of truth. This is sometimes ok.
2. Estimating Short-Term Productivity Decline from Sleep Deprivation
As someone who cares a lot about my productivity, having a rough number for how much less productive I will be if I were sleep deprived could be a useful exercise so I don’t fall into a planning fallacy of believing that the next day could be productive.
Note that these are extremely rough estimates. If you know of better estimates than I use here, please leave a comment and I can update these very rough calculations.
Let’s say
25% of the time when I am sleep deprived
- I listen to 1 hour less podcasts/audiobooks while doing chores like cooking or taking walks because I am too tired to listen to them.
- I am 30% less productive when I am doing work. This means that if I were working 8 hours, I would take 2.4 hours longer to complete tasks.
- I spend 1.5 additional hours unproductively feeling awful.
This would amount to approximately 1 + 2.4 + 1.5 = 4.9 less productive hours in the day.
50% of the time when I am sleep deprived
- I listen to 30 mins less podcasts/audiobooks while doing chores like cooking or taking walks because I am too tired to listen to them.
- I am 10% less productive when I do work. This means that if I were working 8 hours, I would take 0.8 hours longer to complete tasks.
- I spend 1 additional hour unproductively feeling awful.
This would amount to approximately 0.5 + 0.8 + 1 = 2.1 less productive hours in the day.
10% of the time when I am sleep deprived
- I am 5% less productive when I do work. This means that if I were working 8 hours, I would take 0.4 hours longer to complete tasks.
This would amount to approximately 0.5 + 0.8 + 1 = 0.4 less productive hours in the day.
Therefore, the expected value estimate is that I will be losing 0.254.9 + 0.502.1 + 0.10*0.4 = 2.315 hours from a lack of sleep a day. My guess is that this is likely to be an underestimate because it doesn’t take into account poor decisions being made (choosing to prioritize task B over task A), the increased likelihood of falling sick if you are sleep deprived, and the increased likelihood of burning out for an indefinite amount of time, and the fact that it takes around 4 days to make up for the lack of sleep. These estimates also do not account for the likely decrease in lifespan that may occur due to a chronic lack of sleep.
That being said, productivity is not the only thing that suffers from the lack of sleep. It also impacts your subjective well-being, desire to exercise, ability to moderate your eating, social interactions, and more.
3. Enumerating Failure Modes
While I generally have an intuitive idea of why I fail to go to sleep early enough, explicitly listing out the failure modes leads to more concrete and effective targeting mechanisms.
Here’s an example
- I realized that I stayed up to finish a task (downloading software and trying it out)
- I list out this failure mode and strategize ways to combat it. I realized that I was excited to get the task done and afraid that I would not ever finish the task if I didn’t stay up to
- I started running a recurring script (for Mac) every night that will play "You might think that you'll never do the task if you don't stay up to do it. If the task really is that important, you will do it later; if the task isn't, then sleep is more important" to my speakers / headphones.
Surprisingly, it worked. This is probably because
- I internalized that sleep was more important than whatever it was I was doing after doing the above 2 exercises
- The script reminded me to start my nightly routine
- The timing of the script corresponds to when my website blocks start
Here is the link to a template for this spreadsheet.
I’ve been doing this consistently for a couple weeks already. I use Tab Snooze to automatically open up the tab every morning (recurring tab openings show up as a pro feature but I haven’t paid for it and it still works for me).
Recap
A copy and paste of the TLDR
Things I’ve Tried Shortened
- Paying someone else $0.01 for every minute later I go to bed than my bedtime. (Perhaps next time I could use SPAR).
- FocusMate while doing bedtime routine.
- Posters reminding myself to sleep.
- App/Website blockers.
- Reading Why We Sleep.
- Making rough calculations on productivity loss.
Things I am Currently Trying
- App/Website blockers.
- Listing out the ways I fail to go to sleep earlier and strategizing ways to combat those failure modes.
Mindset Shifts
- Framing staying up late to meet deadlines as a high-interest loan.
- Recognizing planning fallacy with variability in how we feel the next day while being sleep deprived.
- Realizing that giving up sleep to be successful backfires.
Recommended Exercises
- Reading Why We Sleep.
- Making rough calculations for loss of productivity on sleep-deprived days.
- Enumerating and combatting failure modes that lead to sleep procrastination.
Helpful Resources
- Mindset Shifts: Why We Sleep
- Website Blockers: Freedom, Cold Turkey, FocusMe, Leechblock (browser only)
- Enumerating failure modes: Spreadsheet template
- Insomnia: Many Great Articles
Acknowledgments
I appreciate Sydney Von Arx, Constantin, Chris Lakin, Raj Thimmiah, Talya, Kevin, Hawk, Ti Guo, Alan Taylor, Tony, and Aaron Gertler for reviewing a draft of this post and providing feedback. I take responsibility for all errors in this document.
I am also grateful for everyone who has nudged me towards prioritizing sleep and sharing strategies on how to do so. I definitely feel better with a more stable sleep schedule.
PS: Say hi
Some of my most interesting and life-changing conversations and insights come from people in the effective altruism community. Even if you don’t feel like you have anything to say, you probably have something interesting to share. All of us come from different backgrounds, so something that is obvious to you may have been overlooked by me.
What does this mean?
- Let’s cowork: I enjoy meeting new people at coworking sessions. I’ve found them to be productive for both participants! Please do not hesitate to schedule one at https://calendly.com/coworkwithemily/60min
- Low barrier to commenting on this post or my anonymous comments form
- Low barrier to checking out my personal site https://emily.fan
- Low barrier to emailing me at emilyfan@mit.edu
- Let’s chat: I am more likely to cancel these meetings than coworking to keep myself from becoming too overwhelmed, but again, I love meeting new people: https://calend.ly/emilyfan/30min
Cross posted on blog.emily.fan and Less Wrong
However, if you commit to going to bed at a reasonable time, and as a result don't end up finishing the presentation, you will suffer in the short term. The next time a similar situation occurs, you will have a very hard incentive to start work earlier, and thus avoid this completely. Timeless decision theory may also be relevant here. ↩︎
The notion of sleep as a high interest loan was inspired by a friend who was inspired by the CFAR handbook. ↩︎
I found these red glasses more intense than the orange ones. Note that they smell bad initially, so you might want to first air them out. Also, this is an Amazon affiliate link if you feel like supporting me with your purchase and are based in the US. Otherwise, this link should work, and no hard feelings. ↩︎
Again, this is an Amazon affiliate link if you feel like supporting me with your purchase and are based in the US. Otherwise, this link should work, and no hard feelings. ↩︎
Another common supplement is melatonin. Melatonin worked great in terms of falling asleep quickly, but I did not wake up feeling well rested. ↩︎
Again, this is an Amazon affiliate link if you feel like supporting me with your purchase and are based in the US. Otherwise, this link should work, and no hard feelings. ↩︎
Another data point: IIRC, Will MacAskill says he gets 9 hours or he doesn't function at his best. (I think I heard that in his Tim Ferriss interview.)
Some things that have helped me to sleep, in case they help you:
Thank you for the thorough post.
I think reading Why We Sleep is good for some people, but may be harmful for others.
If you are already convinced that sleep is really important, but find it hard to fall asleep, and therefore going to bed starts feeling aversive, then reading a lot about not sleeping enough will harm you will probably only increase anxiety, making it harder to fall asleep.
In that case, I recommend reading Say Goodnight to Insomnia by Greg D. Jacobs instead. I read Why We Sleep first, and then Say Goodnight to Insomnia, and it is remarkable how different the two books interpret findings on sleep. Jacobs basically says that it's really normal to sleep less than 8 hours and it will probably not harm most people much. I found it really helpful to become more relaxed about sleeping, and it also has a lot of actionable advice.
I love this post. It singles out a very specific problem and tackles it very thoughtfully.
On website blockers: I have also quitted them regularly but since I have started using ColdTurkey I have quitted much less. I think it's better than other blockers.
For myself, family life has done the trick of making me go to bed at a reasonable hour. But as soon as my wife and kids are gone for a day or two, I (regrettably!) just stay up forever. One of the reasons why I do so (and which doesn't come up on your list) is that my mood often happens to be very good when I stay up late and I also enter flow states more easily when working late at night.
Nice post; awesome illustrations.
I really love the https://heyfocus.com/ site blocker (bonus: it tracks how much time I spend working through the day, and I can easily see my work progress in chart form by day, week, month, year).
Got that tip from this article: https://markmanson.net/attention-diet, which also lists other good site blockers, and tips for focus. (skip down to "Step 1" onward, and especially "Step 4".)
Thank you for the nicely illustrated post! That's what I had been doing for years - reducing my sleeping hours to prolong the time I could dedicate to work or study. I had to face the consequences later when my health issues started. Numerous researches show that a human organism can function well only when having enough sleep - that is, 8 hours for an adult. I stumbled upon an interesting research paper https://ivypanda.com/essays/the-eight-hour-dilemma-sleeping-time-reduction-when-a-single-hour-makes-a-difference-research-paper/ with an idea about the quality of sleep over quantity. While the eight-hour regime is considered optimal, it can no longer be the ultimate amount of time for a healthy sleeping regime. The author states: "...while ones might need eight hours of sleep, for others, seven hours can suffice, and some might demand nine hours according to their own bodily functions and features..." As for me, I still think eight hours of sleep are optimal, and I feel best when I manage to reach this amount of sleep nightly. Thanks again for the deep insight into the topic and the recommendations on stopping sleep procrastination.
Thanks a lot for the thorough post Emily! I like the framing of staying up late as a high-interest loan a lot. And I agree that reading Why We Sleep may indeed be quite useful for certain people, despite its shortcomings. You make a lot of good points and provide several interesting ideas, plus the post is written in a very readable way, and the drawings are great.
Not that much else to add, except two tiny nitpicks regarding your estimation:
Some things I do:
Mandatory link: https://guzey.com/books/why-we-sleep/
(Author of the blog post is a good friend of mine, which makes me biased. I am very unlikely to engage in a discussion of the book's or post's pitfalls and merits here.)
The author of this forum post already links to that :)