Hide table of contents


It went okay. The office increased EA-related interactions and allowed for fellowship discussions to go on longer than they otherwise would have, but managing the office was costlier (in time and money) than the positive outcomes. Yale EA leadership decided not to continue with the office for the next year.  

 Thanks to Thomas Woodside, Alexa Pan, Jessica McCurdy, and Eui Young Kim for providing feedback.


Intended audience: People who think about community building. 

The motivation of this post is to describe the outcomes from the Yale EA office and to give some context to people who either manage EA offices or are thinking about creating an office. 

This post will contain some commentary on why we created an office, how we used the office, and some things I’ve learned that I want to share.

Theory of change

When EAs talk and hang out with other EAs, good things happen. Conversations are very good at getting people to think harder about EA things.  I think that “thinking hard about EA things” cashes out into students working on the most pressing problems.  But going from “interested in EA” to “dedicating their career to work on whatever is most impactful” is better understood as a complex system rather than a step-by-step linear process. Having conversations is only one component in a much larger thing, but it’s an important component nevertheless.  

 The primary way that EA student groups get conversations to happen is through one-on-one conversations and fellowship discussion groups. Each of these comes with costs to community builders’ time. 

One-on-ones require some central planning. The responsibility is on community builders to identify who should talk to who. 

The office would make this less bottlenecky by creating a space where people figured out who they wanted to talk to on their own. This is both more organic and more conducive to people becoming friends.  

Fellowship discussions are another place that people go to talk. Sometimes, when things were really great, people would want to stay after the fellowship to talk more. Because we had full control of the office, those conversations could continue without being interrupted by the next student group that had the room reserved.

Having a dedicated space for people to work together can make people feel more involved, expose them to more spontaneous information about opportunities, cause them to have more conversations with others who are very committed, and otherwise increase the ambient amount of EA-relevant interactions that occur in the student group. 

Some universities have math lounges. It’s like most other places on campus except with more chalkboards and near math department offices.  Math lounges are the default location where math collaboration happens. Essentially we were trying to create a math lounge, but for EA.   

Given the potential benefits, the Yale EA leadership decided that it would be worth experimenting with an office space. We created a doc that expressed this to members of YEA.  It is viewable here.


Office search considerations

There were a few considerations in deciding where to have the office:

Minimizing transactional overheads. 

  • I thought of the office as a minimum viable product to test the hypothesis that a common space would increase the metrics we hoped to increase.
  • We wanted a system that was easy to deal with.  I didn’t want to have some transactional overhead where I had to deal with a landlord and be especially careful with which paperwork I signed.  I wanted to work with an agency that makes it easy to lease and to end the lease. In case the experiment failed, we didn’t want to be beholden to a lease that extended past when the space would be useful. (e.g. a semester is long enough to know whether or not the current space is working, so we didn’t want to sign a lease for a year)
  • Outcome: This went well, the company was very easy to communicate with, and their payment portal was easy to use.

Distance to campus 

  • Students don’t like walking.  Our EA group house was roughly 14 minutes from the center of campus, and the distance has been a source of negative feedback for our group’s social events. We looked for office spaces that were close.
  • Outcome: Our office was eight minutes from the center of campus. Also, the math lounge was also eight minutes away from the center of campus. However, the location of the office had a few marks against it: 
    1) The office was off campus, which made it feel psychologically further than it was. 
    2) Relatedly, the walk to the office was eight minutes in an inconvenient direction.  i.e.  Even if it was only eight minutes from the center of campus, the office was something you had to deliberately walk to. 
    3) The office was located near an area where some students didn’t feel safe walking around, especially at night.


Avoid unnecessary costs

  • If we didn’t want to continue the office, we didn’t want to have a bunch of extra furniture with nowhere to put it. We preferred a furnished office.
  • Outcome: The office came with good office chairs and desks.

Easy access control

  • It was important that it was easy to get in and out of the office without trouble.  We looked for offices that made this easy.
  • Outcome: The building security made access control more difficult than it needed to be. I spoke only to the floor we rented from about letting up guests and members. In retrospect, I should have spoken to every party (e.g., the building security guards) responsible for getting people from outside the building inside.

Office outcomes

How our members used the space

  • Yale EA leadership meetings
    • The leadership met at the office. In one meeting we had an especially productive debate on the overall strategy of the group. Because our meetings were held off-campus, we were able to continue after the meeting was over.
  • Open-invite discussion dinners
    • These were well-attended. I suspect that we would have even more people attend if they were not held at the office. It was uncommon for new people to stay after dinner.
  • In-Depth EA (IDEA) Fellowship & Cold Takes discussion groups
    • Some groups stayed after to continue the discussion. Again, because we had our own space, members didn’t have to clear the room immediately after for the next student group.
  • A few one-off workshops
    • I would guess that attendance was lower than it would have been if workshops were hosted on campus.
  • General study space
    • This was the main appeal of the office. We created a schedule to accommodate different study preferences.  This is what it looked like: 
Note: Two IDEA fellowship groups and the two Cold Takes reading groups met in a shared room on the same floor as the YEA office so that others could use the office as they pleased.  

General thoughts on how people actually used the space

  • The number of people that used the office was significantly lower than what I had hoped. Only a select few people went out of their way to regularly use the space. Over the semester, they became friends, but they weren’t particularly focused on the ideas.
  • In general, I was quite disappointed with the way that the office was not a space for rigorous discussion. Every once in a while, there would be some lively discussion on some EA topic, but that was the exception and not the norm.

Lessons learned 

  • An office should be more central to campus.  Even more central than eight minutes from the center.
  • An office should be close to where students hang out.  Even though it was only eight minutes from the center, it wasn’t in an area of town that students would frequently visit.
  • Preferably the office should be a place all to ourselves.
    • The office was located in a big building with a bunch of companies. It wasn’t a particularly happy or motivating office floor.
  • The room had a somewhat scrappy aesthetic. If I had to do it again, I would be more willing to spend money on items that increased the coziness of the space (soft rugs, blankets, better lighting, etc. ).
  • An office should have at least two rooms. Make sure that it has more than one mode.
    • Our shared space was just one not very large room. This made it difficult to have more than one thing going on
  • Make access control easy for the people that you want to be there.
  • Have some sort of system that makes it, so people know that people are there.  It’s no fun working alone, especially if it’s out of the way.
    • We tried doing this by having a weekly schedule
    • Another thing we did to try to achieve this goal is to have food.  For example, I would bring bagels on Sunday mornings and announce it in the slack.
    • We tried to have at least one exec person be there at all times
  • Some of us were worried about the optics of a student group supposedly dedicated to altruism spending money on getting an office. Fortunately, it seems like this info mostly stayed internal to the group and didn’t have an adverse effect on our reputation on campus. That said, it’s important to provide your reasoning for why you think it’s worth the monetary costs.
  • The people who used the office the most aren’t necessarily the people who are most involved with or serious about EA. Some people came mostly to co-work or hang out with friends, but this didn’t lead to further engagement. Other people who came would have been really engaged anyway even without the office. This isn’t necessarily a problem but does suggest that you need to be intentional about how people use/think about the office (e.g. what is the value of co-working time?)
  • Snacks and drinks make people happy. Fresh fruit is good.

Really basic cost-benefit analysis

Time costs: Managing the office was cumbersome. I spent about 4 hours per week directly working on it.  The other office manager spent about 1 hour per week, with more time during the beginning of the semester.

Monetary costs: We spent about 2400 on non-rent costs in total and about 2500 per month on rent. 

I am fairly confident that I would better achieve the goals that the office was intended to achieve by spending my time on something other than managing the office. For example, one-on-one conversations and creating a forecasting group both feel like they would increase the overall level of engagement and feeling of belonging better on an hour-per-hour basis than managing the office. 

When considering whether to continue with a shared space, we decided against it. The value that we got from the office was not worth the time and monetary costs.





More posts like this

Sorted by Click to highlight new comments since: Today at 10:54 AM

Thanks for this post. It’s good to see that spending on community building is being thought about carefully and benefits are being weighed against costs.

Of course feel free not to share, but I'd be curious for a photo of the inside of the office! Partly I am curious because I imagine how nice of a place it is (and e.g. whether there is a fridge) could make a big difference re: how much people tend to hang out there.

There was a fridge and the snacks were laid out very nicely (though I don't have any pictures, maybe somebody else can share). I personally thought that our room was a pretty nice space, though the hallway/entryway was weird and kind of depressing. The biggest thing that made it nice for me was the view:

Here's a photo during one of our discussion dinners:


On the left, we had a mini fridge, a table with various snacks, a coffee maker, and a water container. 

I think this is a great post and I'm really glad that you wrote this up. You were very smart in the way that you approached this experiment and it sounds like you learned some valuable lessons. In general community spaces are very tricky to establish, and successful ones have to get a lot of details right.

Thanks for this! I'm trying to help setup a new office space, and appreciated you sharing the lessons you learned from your experience. 

Your comment that being off-campus made it feel psychologically farther matches my experience at Princeton. In college I felt like it was way easier to convince people to go somewhere on-campus (except for getting non-engineers to go to the engineering quad) than it was to convince people to go off-campus for something. That's going to be very school-specific though, you wouldn't have the some problem at NYU.