Explore a better way to work – one that promises more calm, clarity, and creativity.

Andrew Gelman’s 4 pm Rule (a Knowledge Work Reverie)

Andrew Gelman is a professor at Columbia University with a joint appointment in the department of statistics and political science. To say he’s productive is an understatement. He’s written six books, has been cited over 120,000 times, and wields an h-index over 100 (if you’re not sure about this last statistic, ask a professor friend to explain why it’s impressive).

The reason I’m mentioning Gelman is a blog post he published earlier this week. As pointed out by the eagle-eyed reader who sent me the article, in the second paragraph, Gelman casually admits: “I never check my email before 4.”

Rationally, this is exactly what you want from a professor at a major university like Columbia: someone who is perhaps not that responsive to emails, but generates six books and 120,000 citations.

And yet, I can tell you from my own experience, and those of the many professors who send questions to my podcast or emailed me in response to my infamous 2019 article on this topic for the Chronicle of Higher Education, this noteworthy habit of Gelman’s is much more the exception than the rule in academia.

An interesting thought to ponder in the days ahead, during which the Christmas break gives us all a temporary inbox reprieve, is what the world of work would be like if Gelman’s email habit was much less exceptional. What if we were all given the cognitive space needed to pursue our field’s equivalent of 120,000 citations?

I’ll have a lot more to say about this idea in the new year, but for now we can set this up as a pleasant thought experiment; a knowledge work equivalent of dancing sugar plum dreams.

22 thoughts on “Andrew Gelman’s 4 pm Rule (a Knowledge Work Reverie)”

  1. Cal you place “productivity” on too high a pedestal, creating an elitist club of members who broadcast to the world how productive they are. Sometimes, being responsive to emails is simply the polite thing to do.

    • I don’t think being responsive once a day is any less polite if the email warrants a response. Constantly monitoring an email inbox so you can immediately respond, even if it is not well thought out or does not actually close the loop, complete a task, or truly move forward a process is no more polite than sending someone an email and simultaneously calling someone to ask “did you see that email I sent you.” It is just one more form of communication, but one that seems to me to be more often a closer relative to snail mail I never asked for rather than more lofty communication like meaningful in person conversation. The kind of conversation that can be hard to have with someone who is watching email pileup on their computer monitor in between you and them.

      • ‘Closing the conversation loop’ is the perfect statement. If email responses are not crafted in a manner that closes the conversation loop, an email chain can go on all day! Cal’s book really helped me understand how to email others optimally to minimize the back and forth.

        I find that I am more productive when I respond to an email when I am in a good headspace. For me, a good headspace is one where I have taken care of some major tasks for the day. I have found that ’emails’ now feel like texts or a real-time conversation. When I first started in education, the expected response time has decreased from 48 hours to 24 hours. I see that people now respond in 24 seconds! Completely unrelated, but I find people are making a lot more mistakes in emails with the decreased response time.

      • Yes that’s taking care of one self and making themself more valuable to others in the long run.

        I’ve also been hearing a fair amount of backlash lately toward those who are doing well or are productive, etc. As if how dare they given the numbers struggling with a pandemic downturn.

    • If it’s that important then email probably isn’t the right medium to try and make contact. Otherwise, the impolite thing to do is to think you’re so important that someone should reply immediately as if their time isn’t at least as valuable as yours. In the knowledge work space, specifically in academia, things would run much more smoothly with people being more self sufficient and thoughtful. Then we wouldn’t have situations like the one where my boss had to send an email to all faculty essentially scolding them for their lack of impulse control and consideration of others in both the volume of email they sent and the arrogance they displayed if they weren’t immediately catered to.

    • This is the typical caveat critique you are giving. Cal’s article is mainly addressing people who want to be more productive but are distracted by email. Of course his advice won’t apply to every circumstance. Moreover, it has become common to immediately reply to somebody. That doesn’t mean that it’s the most optimal way. Fish don’t know they’re in water.

    • Corey, your comment brings up an interesting point: There is no agreed upon timeframe for when someone needs to respond to an e-mail. With snail mail, there was no pressure to respond immediately, and ditto for the early e-mail days when access to a computer and internet connection were limited. However, now that we have constant access to e-mail, some expect an immediate response and are offended if it isn’t received. Unfortunately, the possibility of offending these people results in many of us spending more time checking e-mail than we otherwise would (although this is perhaps becoming less relevant with the increasing adoption of Slack).

      I view Cal’s writings on this topic as constructive, because hopefully individuals who are offended by slow response times may begin to realize (after being exposed to alternative viewpoints like Cal’s) that we aren’t slow responders because we aren’t polite or because we’re elitist, but rather we are slow responders because we feel that e-mail is not an efficient platform for urgent communication and that it distracts us from our primary work objectives.

    • It depends on the job.

      As a project manager at an environmental consultant firm, my job is to deal with problems that arise, whether they be from the client, my workers, other managers, etc. My job is to make decisions and to direct people to enact those choices. As such it’s not just polite, but often vitally important that I be responsive to email. It’s the primary way to carry out these discussions, as it provides a paper trail (it serves as a good litmus test–if you’re not willing to put your idea in writing, it’s probably not a good idea, to put it mildly).

      My closest colleague (my wife calls her my work wife) works differently. She collects data, synthesizes it, and presents recommendations–report writing, in other words. She routinely disconnects from email for a day or two, because that’s what it takes to get her work done. Similarly, my field staff–the folks out sampling wells or drilling holes–can’t be on email, it’s a safety risk (look at the number of construction managers run over since the rise of the smartphone). In both cases, they check email after everyone else is home for the day. Nothing on their plate needs to be done today.

      Am I less productive than my colleagues? Not at all–it’s just a different role. It’s not my job to write hundreds of articles; it’s my job to coordinate efforts between a huge number of people and to draft a path forward on a few million dollars worth of remedial efforts. So I need to work differently.

      I think ultimately that’s the issue here: What is your primary job? If it’s coordination and management, you need to communicate a lot. If your job is to write, you need to avoid communication more. Regardless, it needs to be a conscious choice. Unfortunately right now people default to the standard, which is constant communication.

  2. Since I uninstalled Instagram from my phone, I was able to produce shortly afterwards insightful work that unblocked a project my company was working on for months. I was also able to regain again my focus and taste for books. Society nowadays really underestimate the negative impacts of being connected and reacheable at all times.

    • A freelancing software developer that spends a better part of his/her day developing a piece of rare and valuable software will do well in the long run than someone that’s squirreling between email, reddit, linkedin and twitter.

    • That’s not the point of the article. He’s not saying that everyone should check emails after 4pm. He’s providing an example of someone who analyzed a problem and found an applicable solution to a scenario. The bottom line – applicable to everyone – is to analyze the problem in our daily life. For example, how about checking the email for 15 mins every two hours at a specific time, replying only to the most important ones and flagging the others for a “batch email session” at the end of the day?

    • I am a freelancer (translator), and I only check my e-mail once a day at most. On many days, I don’t check it at all.
      Because besides being a freelancer, I am a human being with lots of other interests. And clients shouldn’t take themselves that important.

  3. If you really need something now is why a telephone is still found to be valuable as too much can be lost in ab email request if its not worded precisely. This wastes both people time, lessens everyone’s productivity, while increases the level of frustration between the two parties.

  4. I love this. I cannot implement the same solution of waiting until after 4pm, but I did turn Outlook’s email desktop notification off, and adjusted my settings so Outlook launches with the Calendar view upon opening in the morning rather than the Inbox view. Then I fit in “check email” on my calendar, usually in the 20 minutes leading up to all meetings (where I know that everyone else will be caught up on the email threads and so not only would I look like a slacker for showing up without being caught up, but I have to have read it to be able to contribute to the conversation and not be lost and confused by what others are talking about).

  5. I know a few professors with incredible publishing stats (definitely not as incredible as Prof. Gelman, but still). A lot of them are so productive, in fact, that they manage to almost entirely do away with the major problem of higher education: actually devoting some time to teaching those pesky little students. By this I don’t mean they don’t lecture – they do their part by narrating the contents of the brilliant book they wrote on the subject for 90 minutes. I mean that they are so focused on producing papers and books, that the average student can expect very little (if any) one-on-one contact with them.

    Among these highly productive professors there appears to be a general consensus that Universities have no social responsibility whatsoever in teaching new generations – only insofar as creating new knowledge is concerned, not transmitting the “old” knowledge on which it stands. I think the following quote, with its blatant omission of teaching as something which is wanted of a professor, illustrates it well:

    “Rationally, this is exactly what you want from a professor at a major university like Columbia: someone who is perhaps not that responsive to emails, but generates six books and 120,000 citations.”

    I hope that it is clear that I am not suggesting that someone who only checks email after 4 PM has no time for his students. I am certainly not suggesting that this is the case with Cal or with anyone around here. What I am saying is that it often feels that the kind of productivity these sorts of posts advocate is solely based on a researcher’s output as a scientist, with no consideration for the other half of the University’s social mission – as illustrated by the quoted statement. This other half cannot be achieved in silence or in isolation. There is a limit to how “productive” a University professor can be, if he/she is to take any part in the social activity of teaching.

    • It’s not just among the professors that teaching is treated as unimportant. It’s a leadership and administration issue. Writing successful grants is a major part of how universities fund themselves. Faculty who can get that money, complete studies, and get papers published are the golden goose, and as long as they teach well enough during their first 4-5 years to get tenure then that’s all that the university cares about. There are still many faculty members that try to balance things and work on their teaching skills, but I can name a handful where I work that are offended by the idea that it’s even their responsibility to know how to set up assignments and enter grades in our LMS, much less actually teach the content well.

      Curtis Bonk (Indiana) and Tom Davis (Georgia) were part of a panel about publishing at the 2019 AECT conference, which focuses on ed tech and instructional design in higher ed, and addressed the issue pretty bluntly. They’re two of the most respected leaders in the field and both admitted that their volume of publishing during their PhD studies was enough to get hired 20+ years ago and wouldn’t even get them looked at by the schools where they work today.

      Until there’s a systemic change in higher education where research and publishing is not 90% of the equation, we’ll continue to see an increase in faculty that view the actual job of teaching as an annoyance to be foisted onto grad assistants or any other plebe they can.

  6. It is inspiring to see that level of success and focus. However, it could be beneficial to know the “rest of the story” as well. My assumption is that Gelman has a junior staff to monitor and answer administrative emails and screen questions from students that can then leave him with only the important emails that can afford to wait. He likely receives and answers a normal level of emails … it’s just that he can afford to do it through subordinates instead. Secondly, based on his age, I am assuming Gelman rose to prominence in his field prior to email becaming the defacto form of communication. He was already successful and established enough to slowplay the implementation of this new tool without effecting his career or relationships.

    It’s a parable-like example for sure, and it is useful for helping us question the utility of our tools before implementing their use. In this example, it is easy to infer that Gelman looked at email during its rise and rejected its adoption as his default communication mode. So we also do with any new tool that comes along. Yet, I’m also beginning to wonder, how many of these examples are just survivorship bias on display. It’s not necessarily the right template for a 20 something just starting out in a culture that expects more frequent email usage. The political and admin costs could easily outweigh the effectiveness of workflow.

    • These are valid concerns. However, I know of at least a handful of examples where people went from “answer emails all day” to “I’m going dark for the next three days to do this important work”.

      Here’s what I’ve found is key to success in those cases:

      1) Have an established name for yourself. If I’m a green-as-grass new hire, I don’t get to tell my boss “I’m not going to communicate for three days”. After a while, though, you get a sense of what’s really important enough to spend the political capital on. I have a colleague who routinely does this, because she’s just that good and the report is that important. No one WANTS to pester her during these times.

      2) Make exceptions. Even if I go silent, there are people I will always drop what I’m doing to talk to. I know that if that coworker I mentioned above calls me, emails me, or IMs me, it’s not just to chat; it’s to resolve some issue that I’m going to have to resolve eventually anyway. Likewise, she knows that if I call her it’s because it’s serious. This allows us to keep tabs on things while we’re not communicating, so that if something major does arise we can deal with it.

      3) Schedule administrative time. It’s not just emails–expense reports, mandatory training, mentoring/staff meetings, and the like pile up as you work on something else. People need to know both that you’ll get to these, and when. “I’ll get to it when I get to it” doesn’t work. Remember, you’re burning political capital by going silent; you need to extend a few olive branches. If you offer a time to do these, it shows that you’re thinking ahead while still allowing you a greater degree of control over your schedule.

      4) Be very, very good at what you do. If my colleague above says she needs a week to work on something, everyone knows it’s going to take a solid week of uninterrupted time–and that the final product is going to be stellar. If you go silent and produce garbage, it doesn’t work. I’ve found that a lot of people think being good enough to not get fired is good enough. That doesn’t work here. You need to be good enough that people are willing to inconvenience themselves in order so that you aren’t inconvenienced (this is why you’re burning political capital by going silent).

  7. As a self-confessed “email” junkie I have got it down to three times a day and never at weekends or holidays. What helped was removing the email apps from my iPhone and iPad. I don’t go into emails first thing in the day, which leaves time for some deep work before I even check email. I have been so much more productive as a result.

    On a personal note, one of Andrew Gelman’s books got me through my PhD on Bayesian Statistics a few years ago, and I have heard him speak at many conferences. A very influential figure in the statistical world.


Leave a Comment