How do you determine whether or not to attend a particular conference? As a working example to see through to natural generalizations, let’s use the JMM. [Just to choose a conference we’ve all heard of, attend at least once in our careers, and a conference that literally has something for everyone. Not to mention a conference that just concluded.]
There are at least two ways to measure expense. The first, and most natural, is money. Starting in graduate school and continuing to my current position, I have had anywhere from $\$1000-2000$ /year to spend on travel. While places will always say there’s “wiggle room” in that budget, an increasing number of colleges are increasingly strict on staying within budget (or even in cutting the budget).
Depending upon date and membership, JMM registration is $\$90-150$ for a graduate student, and $\$360-727$ for someone beyond. Let’s go with the respective averages of $\$120$ and $\$543.5$ per badge.
Then there’s airfare, which most participants will need. When the location is central to many colleges (e.g., Boston 2012) or is one with a direct flight from anywhere (e.g., Atlanta 2017) this number will be lower. But this year, it’s Denver. From my location, that’s about a $\$250$ flight (looking in early November). In addition to the flight, you have to get to and from the airport on either end (parking? public transportation (not always available in college towns)? Uber? Rental?).
Next is the hotel. The JMM is always a Wednesday thru Saturday/Sunday. Supposing you want to arrive Tuesday and leave Sunday, that’s 5 nights. As of early November, 3 of 10 recommended hotels are completely sold out and an additional two are out of student rooms. Suppose you double-up with a colleague or fellow student to save costs. Your contribution to the bill is still around $\$400$ total. You can be clever and book an AirBnB. But that’s still at LEAST $\$200$ for that many nights, and now you get to deal with finding your way around Denver via public transportation or “worse” every day/night.
Last: most colleges don’t take “per diem” into account, but if yours does (like my institution), you can tack on another $\$350$ or more to the bill.
So rough/conservative estimate: to go to the JMM, a grad student needs $\$750$ and someone else $\$1100$.
And you’ve just spent most if not all of your travel budget for the entire year on ONE conference. Because trust me, the bigger the conference, the less likely you are to have anything subsidized. Bill is all yours.
My current position asks us up-front to list for our entire year all the places we’d like to go (with “justification” including giving a talk, etc.) and to estimate the cost. This is different from every other job I’ve had. At other schools, I’d think to myself “How am I going to spend that money?” or “How much do I have left?” while planning my trips, but actually writing it down and budgeting–let alone justifying the trip to highers up–is an extremely helpful activity. It helps you see just how much money it takes to go anywhere, and really makes you prioritize your trips. It also might encourage you to apply for travel grants.
Still, there’s another way to measure expense: time. The last few years the JMM has started in the first or second week of classes. This year is no exception; looking at all the places where I’ve worked, the spring start dates are January 4 thru January 13th with the JMM beginning the 14th. Something I didn’t have to worry about in grad school but have ever since: substitutes. And non-trivial subs too (it’s not like the first or second week of class you’re giving a midterm and all you’d need is a proctor, or that you’re always going to be teaching calculus). Some of my colleagues could miss 9(!!!) class meetings for this trip. And at smaller schools (e.g., my postdoc institution of Davidson), having half the department go to the same (national) conference during the work week can mean either mass cancelling of classes (possibly with later rescheduling?) or some very busy people back home.
So when deciding whether or not to go on a trip, I do encourage you to think about how many subs you’ll need, when you need them with respect to the term, whether the courses you’re teaching are friendly to sub (e.g., calculus) or are more technical (e.g., real analysis), if other people will be around, who and how many you could “owe” later, etc. Some people, of course, take the opposite approach and want conferences that conflict with classes. The alternative, after all, is that the conference conflicts with “me time.” To each his own.
This is something I first internalized thanks to my adviser (who is dead–Erdos sense only–but still it perhaps should be taken with a grain of salt). He told me the JMM was a giant ant hill where every ant had a favorite crumb. And these ants would run around the hill all day and feel eternally grateful to the ant gods for having as little as 5 minutes to convince other ants their favorite crumb was worthwhile. And maybe for a few minutes afterwards, the ant audience would appreciate the ant speaker’s crumb and be really excited. But also just as likely, the ant audience would return to their own crumbs and the net effect of crumb exchange would be minimal.
While bleak and mildly bizarre in choice of ants and crumbs, there was some truth to what he said. Most talks at the JMM are under 20 minutes. With a research talk, the more technical the subject, the harder it is to get ANYTHING out of it if you’re not already “in the area.” But suppose you do get something out of it, you have questions, or just want to follow up with the speaker. Think ahead to when exactly that would happen. A lot of people at the JMM literally run around the conference center like they’re being chased by a bear. It’s hard to find time (or restaurant space) to have a (quiet) conversation about someone’s work or even your own. It can be challenging finding a relatively unpacked corner of the center–and you know you’re never going to find a place with a black/white board! Sure, people end up doing math in hotel lobbies on note pads, but that seems only mildly better than the oft-criticized interview in the hotel room.
Of course, the obvious problem with my adviser’s analogy is that not every conference—JMM especially—is all about research talks. The JMM includes all of the following but imagine what subset might be present at other conferences: There are talks on teaching. Talks on math and society. Panels on professionalism ranging from how to apply for grants to how to prepare a tenure dossier. There are textbook vendors and software developers. There are summer internship advertisers ranging from Facebook to the NSA; graduate program booths and open houses. There are performances of mimes (Tim Chartier), musicians (Lilian Pierce), mathemagicians (Art Benjamin); there are movie showings (Ken Ono), and even an annual game show (Who Wants to Be a Mathematician). There are grad school reunions. Undergrad school reunions. Summer math program reunions. Oh, and then there’s the ever popular employment center.
This type of ant hill environment definitely works for some mathematicians. And some mathematicians specifically like the mix of research with non-research and feel like they can really pick and choose at these big national conferences. But don’t think about anyone other than yourself (and possibly the students you’re leaving with some sub) when choosing a conference. What works for other people may not work for you. You may get better conversations and really get to “meet” people in your area (or interested area) by going small and/or local. In graduate school at UGA and as a postdoc at Davidson, I went to almost every local weekend number theory conference, many of which covered some (if not all) expenses. While I also would “squeeze in” a more expensive regional and national conference if I could for the variety of both topics and speakers, it was so nice to have a literal community of people in my area at every stage (tenured, postdoc, etc.). Yes, the talks were still mostly 20 minutes long, but you could actually find the time to talk to the speakers afterwards; you also didn’t have to worry about badge prices, massive travel bills, or substitutes. These kinds of networks definitely can be harder to find in other geographical areas but if/when you can, it’s magical.
With respect to the JMM, what will be interesting is how it will change in the coming years as a result of the AMS taking over management. The management shift coupled with colleges pinching pennies could lead to an impact on the employment center—an opportunity (for lack of a better word) basically not offered at any other conference. Who knows—maybe online videochats will become preferred as budgets tighten? But as long as that employment center is around, and to those who know they will be on the market: I HIGHLY recommend not having your first JMM be the one when you’re on the market. It is an ant-hill unlike any you’ve ever seen before. You don’t want that to add to the shock, awe, and nerves that being on the market already adds. THAT is worth the money and time.