So you want to go on the job market

2022/03/21

I have spent the last 9(ish) months navigating this thing called the “economics job market.” Really, this just means I was preparing job application materials, submitting applications, and interviewing or (more frequently) sitting and thinking about preparing job application materials, submitting applications, and interviewing.

I am very happy to report that this process has concluded and I’ve accepted a job! On reflection, I’m in a very different place than I was when the job market started. Literally, my desk is now in our bedroom instead of our living room.1 More substantively, I’ve also reclaimed my weekends and evenings, chunks of the week that were sometimes swallowed up by an ever-expanding job market to-do list (or, worse, were disrupted by stubborn internal monologuing about said to-do list). I also now have A LOT more info about this mythical process of which I had only heard department lore over the prior 5 years.

Some of this info is qualitative and some is quantitative. It comes from some combination of: lived experience, observation, and the explicit collection of data/resources over the past academic year. While the spreadsheets, links, and hot (eh, probably tepid) takes are all fresh in my mind I figured I’d bottle them up into something that could be useful for future cohorts.

Consider this job market content part data dive, part advice, part cathartic release.2

Overview of the econ PhD market + broad thoughts

There are a few “job markets” you can go on as an economics PhD student. I mainly applied to government jobs, academic jobs, and non-academic research jobs. If you know you want to limit your search to a specific subset of jobs, you can implement specific strategies for that search. If, like me, you were not down to commit to one path ex ante, you can apply more broadly. From my experience, I’d characterize the necessary steps for a broad search as follows:

  1. Write the job market paper
  2. Set up a job market website
  3. Collect job posting information
  4. Prep your job app materials
  5. Apply to jobs
  6. Prep + do first round interviews
  7. Prep + do flyouts
  8. Navigate offers + take a job

I list the above steps in approximate chronological order but really most are continuous processes that spill into later stage steps. For instance, there might be new job postings that appear (step 3) after you’ve already started applying to jobs (step 5) or you might get a flyout offer (step 7) after you’ve already accepted a job (step 8). The continuous nature of things can make the job market tough if you’re conceptualizing the search as X% finished at any given point. Rather than thinking about your market progress bar, I think it can be helpful to just document/celebrate “inputs” regardless of outcome. I became very bought into giving myself carrots for hitting goals. (E.g., you applied to 10 jobs? Congrats, self, you get an afternoon bakery visit!)

In terms of other broad takes, I found it incredibly helpful (logistically, emotionally, all the “ly”-s) to check in regularly with supportive job market buddies throughout the process. You can keep each other accountable and organized as well as supported and hyped up when need be. Examples of actions your mini-JM (job market) community can take to help one another out: give feedback on spiels and written statements, proofread websites, discuss draft follow-up emails to potential employers, strategize about interview timelines, amp each other up during long flyout days, bond/vent about the horrendous UI you encounter during the applications stage, etc. This all to say, treat the market like a team sport. Helping other people can be very validating on days when you’re having trouble with your own applications/interviews. Investing in a little community like this means you can genuinely revel in other people’s success. There’s a lot of rejection on the market, so maximizing the number of times you get to celebrate a win can be huge. And even if most employers ghost you, your group chat never will.

In the following sections, I’ll go through some thoughts/resources on the JM step-by-step (starting with (1) write the job market paper and ending with (8) navigate offers + take a job). Of course, like with most guides and resources, this is based on my experience and some tips might not jibe with everyone.3 I am one person in one program in one year of the market. With that caveat, I still think many of the things I talk about are universal to the job market experience. I hope this write-up can be informative/useful to a wide variety of people navigating the process.

(1) Write the job market paper

OK, writing the job market paper is less one step than it is the accumulation of many, many steps that came before. It is definitely the most crucial part of your application packet and deserves most of your attention until the Fall. This was the step I struggled the most with, so I definitely have not cracked the code on this. (Who has? Please show yourself if so.) But I do think I made some mistakes I would tell my 2020-2021-self to avoid. (I would also tell her to pack more clothes before her flight to California in March 2020 because she’s not leaving for a while.)

If I could do it again, (this might be controversial) I would take back the times I worked on my JMP draft writing too early. Why? Because the method details and paper focus changed over time, which meant I had to painfully part with weeks of labor I put into writing and structuring that was now outdated and needed to be scraped. In fact, I found that my work on early drafts actually disincentivized me from adjusting based on new feedback because I felt like I was “in too deep” with an approach or structure. Instead of putting time into nicely writing drafts early on, consider drafting the core components of your paper in slide or in outline form. (You could even practice the outline in twitter thread form. The character limit mandates being concise and direct.) You might wonder, “well, then how will I know it’s time to write up a full draft if it feels like the work is never finished?” Er, that’s a great question and I am still struggling to figure this out myself.4 However, the external deadlines of the market will mean that at some point, you will NEED to commit to JUST WRITING, so set a deadline for starting that process and stick to it.

Once you’re at the writing and editing stage (once you’ve settled on your argument and the key components), I suggest reading Larry Katz’s and Claudia Goldin’s “The Ten Most Important Rules of Writing Your Job Market Paper”. Definitely send your abstract and intro around to advisors or colleagues for feedback/comments. (Google docs can make it easy for people to leave comments/suggestions. Readers can even comment on each other’s comments. E.g., B writes, “I agree with what A said here.”) Avoid iterating after each individual person’s advice. Instead, try and get a bundle of takes (a portfolio of suggestions) and then assess them together. It might became painfully (but usefully!) clear that Section X makes no sense because all 4 people who read it wrote a big “?” next to it. But, on the other side of the unanimous-split-opinions spectrum, it could be that reasonable people disagree… E.g., one person thinks the intro is better if you say A before B, but another person thinks the opposite. If you started with B before A and then switch to A before B and then you get advice to switch back you might end up stuck in an endless loop. It is not possible to take every suggestion, so weigh them in the aggregate with your own opinions as well.

While I like the 10 JMP rules write-up, I am going to have to push back on “Rule #9: Be your own worst enemy.” I found that I leaned a little too hard into the “be your own worst enemy” mindset at the start and and lost confidence and motivation. I think it’s worth doing some self-reflection and figuring out what motivates you before the market starts. Sure, you want to have a critical eye for your work but not so much that you lose sight of your skills/talent or just fundamental value as a person. “Your JMP quality is orthogonal to the quality of your person” or, a more human way to say this, “you’re not your JMP” (-John DiNardo). In my opinion, the process itself can tear you down, so it’s worth building yourself up in preparation. The mental element of the job market is immensely important and could use its own specific guide… In fact, Chase Eck, Kelli Marquardt, and Dana Shaat’s wrote up a guide on exactly this topic: A Guide to Mentally Preparing for the Job Market. It’s excellent–definitely check it out.

(2) Set up a job market website

You need a website for the job market. I think it’s worth keeping it really simple and skim-able! People are going to get a quick sense of you in ~1 minute so don’t overwhelm with information.

What do I base this 1 minute estimate on? Why, I’m so glad you ask. While all my potential employers were collecting data on me and visiting my website, I was collecting data on them. I set up google analytics for my job market website, which meant I was able to passively collect information on visitor behavior. This is one of a few data collection exercises I engaged in during the job market, knowing that at the end of the process (that day is today, my friends) I’d get to analyze the final products! This is what we in the research world call “playing the long game.”

Collecting data on my own experiences was also weirdly empowering because it flipped the power dynamic of the market on its head. Sure, I was interviewing with employers trying to get a job, but I was also collecting data about their actions as a researcher as well. I also joked with friends that collecting data on my own experiences helped me disconnect emotionally from the reality of rejections. (That’s healthy, right?)

Anyway, in short, people spent about 1 minute (well, 55 seconds on average, to be exact) on my website.5 Of course, I don’t know the full distribution of average visit duration for all job market candidates,6 but I think we can all agree there are lots of candidates and not a lot of time, so people are not going to be spending a long time on each website.

To keep the website simple, it seems like a good model is to have a home page with: name, headshot, email, fields/research interests, and a statement that you’re on the job market. You can then have obvious page links/tabs for: CV PDF and a research page with links to papers and abstracts (make it clear which one is your job market paper!). You can also include a page with teaching information, but, in general, people tell you that teaching information is much less crucial than research information. That seems to check out in the data… my research page got ~6 times as many clicks as my teaching page during the market. Even a link to my personal website (where you are now) got twice as many clicks as my teaching page.

When should you have a website up? Well, definitely before your department blasts out their job market list. But, I’d argue, all else equal, it’s worth having it by early October based on collected data. Below I show you the number of active users from September 2021 to March 2022 to my job market website.

Active users for my job market website, 9/1/21-3/3/22

Our department made the list of job market candidates live November 1, which lines up with a big spike in active users. However, there are also little spikes in early and mid-October. My guess here is that other departments have earlier dates they post their candidates and that might send people in Econ-land (search committees or others) off to check out people in other departments. The number of actives users per 28 days doubled from September (~200) to the first half of October (~400). This number peaked in mid-November around ~1500 active users in a 28 day period.

How do people find your website? For me, most of my users (64.5%) came from people clicking a link to me from the Harvard econ department page (34.5%) or clicking a link from Twitter (30.9%).7 22.6% of users got there from a google search and 9.6% already had the link. I would assume that 9.6% were people who had the URL through the CVs I sent out for job applications. (This theory checks out in that a much higher percentage of users come “directly” (23.8%) during the period of time places were likely looking through job applications. That is, Nov 15-Dec 1.)

(3) Collect job posting information

There are a variety of job posting websites for people on the econ market. Our department told us about 10: JOE, EJM, European Economic Association, APPAM, job bank, Chronicle of Higher Education, American Finance Association, Social Science Research Network, Royal Economic Society, and Academy of Management.

Before the market, I thought I’d have to scan all 10 regularly to stay up to date on jobs. But, in the end, I found 83% of my postings through JOE and the remaining 17% were from LinkedIn or Twitter. The vast majority of the jobs I saw on the other 9 non-JOE job board sites were already on JOE, so I think at some point (tired of all the extra effort for little payoff), I started only checking JOE and left behind the other 9 sites. (However, I will note that one friend mentioned to me that they regret only checking JOE because there were relevant jobs on EJM they missed.)8

You’ll become very familiar with JOE during this process. Jobs are not posted all at once. They are posted gradually on JOE, so it’s worth knowing what that pacing looks like for planning purposes. Thankfully, Andrew Lilley made some nice plots on this using data from 2019-2020, 2020-2021, and part of 2021-2022. See below for screencaps of these.

JOE listings plots by Andrew Lilley

Of course, there are very interesting things to notice here about that first Covid-19 year (2020-2021), but there are also some generalizable takeaways from a planning and strategizing perspective. Let’s focus on non-academic full-time and Assistant Professor tenure-track jobs (in blue boxes). For these two, most jobs post get posted between September and November. (More get posted Nov-January but the rate of posting majorly slows down.) By early November, you’ve seen about 75% of all the jobs that will posted by January.

How does one keep track of all the jobs and deadlines? There are a few methods here… the basic way is you can check the JOE page every few days9 and keep your own spreadsheet. (Make sure to click into JOE ads for details because the deadlines JOE shows are often wrong.) Or, fancier, you could try out Benjamin Vatter’s “Job Market Application Tracker”.

When are the deadlines? In my sample of job applications (that included deadlines), 60% had a deadline in November and November 15 was the most popular job application deadline date. Here is a histogram of the deadlines I faced to give you a sense of the distribution:

histogram of deadlines

These are the posted deadlines, but conventional (hidden curriculum) wisdom is to get everything submitted before Thanksgiving even if a deadline is in December.10

(4) Prep your job app materials

The most common types of documents you need for job applications are as follows: CV, job market paper, cover letters (job-specific), statements (research statement, teaching statement, DEI statement), and recommendation letters.11

On cover letters: For non-academic jobs, I really do think cover letter quality matters. I heard this from employers before I applied and I experienced people referencing my cover letter details in interviews and flyouts. How do you write a good non-academic cover letter? I used great and actionable advice on writing non-academic cover letters here.

On writing statements: I highly recommend sharing drafts of these with your JM buddies. You can leave comments for each other and discuss. Giving feedback on other people’s statements will help get you out of your own head. (Having group deadlines also helps move drafting these along.) It’s also worth reading statements from program alumni and alumni friends to get a sense of the variety in formatting/style for these. For the DEI statements, UC Berkeley has some nice resources here and here.

On recommendation letters (and managing your letter submissions): You’ll need to get your references (3-4 committee members) to submit letters to many of these places you’re applying. In my experience, there are 3 ways this can go:

  1. The letter is submitted to a general portal (Interfolio, AJO, EJM, AEA JOE) and then you can apply the same letter to all the jobs you apply to through those systems (so easy–love it!)
  2. The employer requires an email submission– this requires you tell your faculty or staff the emailing instructions for each unique place (and then confirming with them that they submitted) (agh)
  3. The employer has a specific online portal that will trigger a request to that person for them to submit a letter (agh also)

If you’re applying to 100+ places, this will be overwhelming. I found it helpful to maintain a spreadsheet with letter status by place (or portal). (Here is my google sheet example for you all.) Stark color coding made it clear to me and to people submitting letters when and where letters were missing. See the below screencap for a way to organize this–there are general portals, email submissions, and then place-specific portals. There are distinct columns for each reference (ref 1, ref 2, and ref 3). References mark the dates they submit. I then made cells red for missing references and highlighted incomplete institution/portal names in red as well.

Users to my harvard scholar page over time

Sure, it’d be great if everyone just used a mainstream portal, but conditional on the current mishmash of options, being organized in some way makes the process easier for you, the candidate, and your support network of faculty or staff.

(5) Apply to jobs

OK, let’s say you have your CV, JMP, etc. and you have a spreadsheet of jobs with deadlines and links… now what? Start applying! This is going to take forever and involve a bunch of admin madness, so… give yourself time!

Of course, easier said than done. I sent in 80% of my job applications in November. See below for the histogram. There was even a day (November 12) were I banged out 22. Would not recommend.

histogram of application submissions

How many jobs do people apply for? How do these translate into later-stage parts of the job market (interviews, flyouts, offers)? The AEA has an informative webinar on the job market that answers some of these questions. In 2020-2021, for rookie candidates in US economics departments, the median self-reported number of job applications was 145. The median numbers of first-round interviews, flyouts, and offers were: 15, 4, and 2, respectively for that population. See Table 1 from the below webinar screencap.

Screencap from the AEA webinar (at 52:56)

The webinar mentions that the flyout and offer numbers are right-censored because you can leave the market before you get the full set of flyouts and offers. That is very true. The market speeds up in January once you’re off to the races on interviews, flyouts, and even offers. The “phases” will overlap, which can be especially challenging if any tight offer deadlines get imposed. The timing is different for everyone, but here is a visualization of how the application/interview/flyout/offer timing played out for me:

showing phases of market

The plot shows the density of application/interview/flyout/offer occurrences over time. The peaks are where occurrences were most concentrated. Applications were spread over the longest interval, but most were in the month of November. 1st round interviews were concentrated in early January (due to AEAs timing). Flyouts were more spread out (which makes sense because you can’t do more than 1 per day). Offers were concentrated in early February. But the offer timing is very dependent on your flyout schedule in the prior phase and the existence of any exploding offers, which might lead to concentration of offers in time. Things get really endogenous and idiosyncratic after flyouts. I share my timing since I found it helpful to have a few examples of how it played out for friends in prior years. But, of course, it is different person by person.

Another way to see these phases is by looking at the cumulative count of each milestone (application/1st rounds/flyouts/offers) over time.

showing phases of market

Again, this shows that phases will overlap and that the market speeds up in the later stages. Another big take-away here, I hope, is that rejection the vast majority of the time is normal and expected. I guess what I’m saying, to quote from Career & Family, is: “The good news? It isn’t you; it is the system. The bad news: it isn’t you; it is the system.”

(6) Prep + do 1st round interviews

Nothing helped me prepare for first round interviews more than some twitter threads Natalia Emanuel wrote. Here is the first one and here is one she wrote about about commonly asked questions.

I would add/highlight a few things about remote interviews:

Perhaps my favorite 1st round interview mantra was pulled right from Luke Stein’s twitter:

I still have an index card with this exact sequence of 3 bullets written on it.

(7) Prep + do flyouts

When it comes to flyouts, I think one of the most important dimensions of this comes before your first flyout even starts. It’s in the scheduling.

Why is flyout scheduling important? In my opinion, because of “exploding offers” or offer deadlines. Some employers might make you an offer but tell you that you have limited time to decide (the most common deadline I heard of was 1 week, but I also heard of 2 weeks or even a few days). If you haven’t done all your flyouts yet, you now have your attention divided between flyouts and thinking about this offer. Moreover, you might not be able to consider all offers at once. Ideally, you’d consider any eventual offers all at once. So, solve the problem backwards.

I think the best way to do this is to schedule the flyouts near one another. (Of course, this might not be possible if you need to physically fly between places… and it also means you’ll be pretty exhausted after back-to-back flyouts.) Regardless of exactly what you do, the point is to try to be strategic in scheduling–take into account how your scheduling might play out in the next phase.

In terms of doing the actual flyout, here’s a great twitter thread on flyouts from Natalia again. I’d add a few additional points:

You might wonder, what happens if you don’t get a flyout from a place that interviewed you? Unfortunately, most of the time places will ghost you. The good news it that you can always reach out and ask for an update. This can be really helpful for getting a better picture of how your option set is evolving. (It’s worth noting that place might come back to you about flyouts later if their first set of picks don’t work out. Things can change really quickly.)

To employers, a few suggestions for remote flyouts:

  1. Please give candidates a long enough lunch break. A 30 minute break during a flyout might not suffice if someone needs to run out to grab a sandwich. I never had lower flyout energy than one day where I kept wistfully looking at a bowl of pasta I had made during break but ran out of time to eat. (Don’t worry–I eventually excused myself for 2 minutes and the pasta quickly disappeared.)

  2. Offer people random breaks for water and going to the bathroom. Build them into the schedule and encourage interviewers to offer them regardless. (Sometimes candidates might lose breaks because they talked through them and didn’t notice the time.)

  3. We JMCs always thought it was super classy when a place bought us lunch/dinner from afar.13 It simulates the feeling of being “hosted” even when everything is remote. Also, recall that grad students love free food.14

(8) Navigate offers + take a job

Throughout the process, you might hear employers tell you to keep them updated. What exactly do they want updates on? My sense is they want to know if you have an offer from somewhere. Once you have an offer on the table, you definitely want to check in with places where you’ve done or are doing flyouts. If you’re facing an exploding offer, definitely let other places you’re interested in know the deadline so you can get a shot at comparing the offers!

For negotiation advice, this lecture on “How to Negotiate Your Job Offer” by Deepak Malhotra is incredible. Honestly, it is just a great video to watch regardless of if you’re on the job market or not. He’s an excellent presenter. Also, here’s a useful set of advice from Claudia Goldin and Larry Katz as well.

Let’s imagine you’re at the point of trying to make a decision about what job to take. This is obviously a super personal decision. To the best of your ability, try and give yourself time to talk to people (family, friends, advisers, peers, people at the different jobs) and sit on it. It’s very hard to do this under tight deadlines, especially if you’re making a joint decision with others in your family.15

Much of the job market interviewing and the job market advice is focused on the job content itself (what is the job day-to-day like?). But, of course, the location, pay, and benefits matter a ton as well. Moreover, some of these components might be new information as of the offer. This is all to say that there are many dimensions to mull over.

When do people accept job offers? The AEA webinar shows that 70.6% of people accept jobs in February (47.3%) or March (23.3%).

Screencap from the AEA webinar (at 59:49)

Off the market

Once you’re on the other side of a process, it’s easy to take your new hard-won knowledge for granted.16 Hopefully, this bundle of recent takes/experiences can provide you with new resources, intel, or tips that can help you during the job market.

Eventually!! you will have a job, you will be off the market… at that point, “food tastes better, the air seems fresher, you have more energy and self-confidence than you ever dreamed of.” Really, though, a plant in our living room started flowering after my job market wrapped up and you cannot convince me that that was not causal.17

As a final goodbye to this process, I’ll leave you with a visual representation of my job market months–one of many graphs I would day dream about creating while I was submitting dozens of applications. 💭

It’s a raster plot that shows how each of my job applications moved (or, more frequently, did not move) through the job market stages.18 If you want the proper, de-codeable data visualization version, here is the same thing but with a legend key and an x-axis showing time.


List of resources

Here are all the resources I referenced or screencap’d at some point in my write-up:


Data/Code

Here is my R notebook with all the code used for generating plots in this post.


  1. For any “post-pans” who ever stumble upon this, these were the things we did when we needed to “change up our work space” in 2021.
  2. Mandatory “content” link.
  3. I wrote this in a way that would have been helpful to my past-self because that’s the younger person I know the best.
  4. This very well might be a Goldilocks situation, in which I now wait too long to write the full draft… will report back.
  5. I’m using google analytics data on my prior JM website from 9/1/22-3/3/22.
  6. Unless you all want to track analytics and share them next year…
  7. I made my website link in Twitter my job market website at some point in the Fall. To the extent I had a different link there, this underestimates the power of twitter to get you clicks.
  8. Ideally, some job posting aggregator would collect posting links from all 10, remove the duplicates, and share that (comprehensive, de-duplicated) output.
  9. You can also follow @EconJobsBot, made by @TheKeveloper, for the JOE job ads in twitter form.
  10. I’ve seen a few twitter threads from selection committee folks saying submitting even earlier is better since your materials might be read before the crowd shows up around deadline time. I don’t know how true this is. All else equal, it’s probably better to submit earlier, but most people are spending that extra time getting their materials prepped and improving their packets. So, as usual, all else is unlikely to be equal.
  11. Originally, I planned to manually collect data on which of these document types were most frequently requested but, listen, at a certain point in November you couldn’t pay me enough to look at my job spreadsheets more…
  12. I got a haircut during December and felt like a new woman.
  13. You can just give candidates a Grubhub gift card, for example.
  14. Flour seminar sandwiches 4ever.
  15. For your paper reading list: “Gender Differences in Job Search and the Earnings Gap: Evidence from Business Majors”
  16. Most of my teaching experience during my PhD has centered around supporting undergrads write theses–a long process that can feel intimidating at the outset… sound familiar?
  17. My “Mostless Harmless” text is audibly groaning from my bookshelf right now.
  18. Now that I stare at it more I realize it definitely has some “light at the end of the tunnel” vibes, which feels exactly right.