About 8% of the time. Definitely not more than 25%. For the brief details, read on.
While compiling data on who was hired for N. American tenure-track asst. professor positions in ecology and allied fields advertised in 2016-17, I noted when a search failed, or apparently failed. Sometimes the chair of the department told me the search failed. Sometimes I inferred it failed because the university put out a press release this fall listing all their new faculty, that didn’t list anyone who could’ve filled the position. Sometimes I inferred it failed because the university is advertising for the same position again this year. As of this writing, out of 235 searches (some of which weren’t filled by ecologists), 19 (8%) failed by these criteria.
Some of those inferences could be wrong. Maybe some new hires haven’t been announced yet because the searches are ongoing, or the new hires delayed their start dates. Conversely, it’s possible I missed some failed searches. There are another 54 positions I checked, for which I couldn’t determine the outcome. There’s no one currently listed on the department website who could possibly be the new hire, and there’s no other reliable information to indicate the search outcome. So the absolute upper bound on the frequency of failed searches in this dataset is 25% (73/289), but I suspect the true number is closer to 8%
I only know why a few of these searches failed. In general, searches fail for various reasons. Anecdotally, the most common is that the position was offered but the offer was declined, often after some period of negotiation. At which point it may be too late to offer the position to someone else, for instance because the other candidates already have accepted other offers. I know of one recent search at an R1 that failed because four different candidates turned down the job.* Other reasons for searches to fail include searches that are cancelled for budgetary reasons, and searches for which no offer is made because none of the candidates were deemed acceptable.
*This does not indicate that there’s anything bad about that job or that department. People decline offers for lots of reasons, the most common being another offer.
UPDATE: out of curiosity, do these results surprise you?
Related posts
Some other recent posts with data on the N. American tenure-track faculty job market in ecology and allied fields:
Newly-hired N. American tenure-track asst. professors of ecology are 59% women
It’s a myth that ads for “quantitative” ecologists are taking over the faculty job market
Dynamic Ecology 
I’ve been on both side of “failed” searches – trying to hire, and turning down a job that then went unfilled. In most cases I’d make the minor quibble that these are not “failed” searches. They are instead just “delayed”. We hire, and get hired, for the long run, and while there are significant costs for everyone associated with searching and interviewing, these are investments in a good match where a new hire will be happy and productive for many years. It’s worth getting right!
And I admit this is a minor wording quibble and doesn’t damage any of your actual points š
Well, just as a ms that gets rejected might get resubmitted as a new ms, I think it’s fair to call a search that doesn’t lead to a hire a failed search, even if its followed by a second successful search for the same position. After all, sometimes if a search doesn’t hire anyone, the dept. doesn’t get to search again for that position. Instead, the dean or provost reallocates that faculty line to a dept. that can fill it.
None of this is to disagree with your point that it’s important to get hires right. In the rare cases when departments decide to redo the search rather than making an offer, it’s usually for good reason, not because they’re incompetent or somehow screwed up the search. I’m just quibbling with your quibbling. š
I buy somewhere between 8%-25% but I definitely think 8% is too low, at least at US R1 universities just based on searches that I know. Maybe 15%?
All the cases I saw were the scenario you described – first (and sometimes second) offers had other offers and moved on leaving things too late in the season to continue (at least a better idea to restart in the fall).
While you’re absolutely right that everybody (including the people that are part of it) call these failed searches, I agree with Steve that it is probably unfortunate phraseology. If the goal is to find good fits, then taking extra time is hardly a failure. Most departments instinctively know this and will quickly opt for a redo rather than go down a path that feels dubious.
“at least at US R1 universities”
Which raises the question of whether search failures are more or less frequent at more research-intensive institutions. Hard to say, since search failures are so rare overall. I’d guess there’s not much if any correlation, but I don’t know.
I’m now wondering if the minority of respondents who were at least a bit surprised by these results expected the true percentage to be lower, or higher. I guess they mostly expected it to be higher, since it couldn’t be much lower? Or maybe not–maybe some people are surprised to learn that any searches fail?
I was a little bit surprised by how high the failure rate is. Given the scarcity of tenure track positions relative to the number of postdocs looking for tenure track positions, Iām surprised that anyone is getting multiple TT job offers such that they have to turn down an offer from an R1 university. Similarly, Iām surprised that second and third choice candidates are gone (presumably employed elsewhere) by the time the first choice turns down an offer. As a PhD student nearing completion, these expectations are probably influenced by my own fears about the academic job market!
If you look in the comment threads on ecoevojobs.net (not something I suggest making a habit of…), you’ll see some people reporting turning down their only TT job offer. It’s rare, but it happens. People might have family reasons for doing so (e.g., spouse or partner who can’t or doesn’t want to move), or might really dislike the dept./uni/city, or have some other reason.
It’s not that surprising that some people get multiple offers. It’s for the same reason that many people who get interviews get more than one: similar institutions that are advertising for similar positions often look for similar things in candidates. If one uni that’s advertising for X thinks highly enough of you to interview you and maybe even offer you a job doing X, odds are decent that others advertising positions doing X will think highly of you too.
And then once an offer is made unis often are prepared to allow candidates multiple weeks or even longer to negotiate. By which time, yeah, they might be left in the cold if their first-choice candidate turns them down.
As Jeremy points out people turn down tenure track offers for many reasons. I personally turned down my only tenure track offer. It was a big research university with satellite campuses. They changed the campus location from the main campus (listed in the add) to one of the fairly large satellite campuses (which had no ecologists)*, and I decided that I’d rather accept a great non-tenure track offer than a tenure track offer that wasn’t a good research fit for me. While I consider myself rather independent, I want to, at least at this point in my life, be at a place where I will run into ecologists/conservation scientists in the hall/lunch room. So many fun conversations happen that way.
I’m surprised it’s that low, just because in ecology, departments seem to interview so few candidates (3-4 seemed quite common where I’ve been).
I wonder if in economics job searches basically never fail. Universities interview so many candidates at the big national meeting, that they, I assume, have a much deeper list of finalists.
“I wonder if in economics job searches basically never fail. ”
Hmm. My understanding is that those are only first-round interviews, and they’re followed by on-campus interviews for the top 3-4 candidates, as in ecology. So those first round interviews at the AEA meeting basically serve the same role as the phone/skype interviews that some (not all) faculty searches in ecology use.
I’m not sure if doing phone/skype/conference interviews before campus interviews cuts the likelihood of search failure. Probably not, I guess? “We decided not to offer the job to anyone we interviewed on campus” is a rare reason for search failure, but seems like the only reason for search failure that you *might* prevent by doing phone/skype/conference interviews first.
I’ve seen several searches run through the 3-4 on campus (either because they had terrible interviews or they took themselves out of the running) go back to the pool and invite more people on campus based on their strong skype interviews. I’ve also seen searches which had run through their on campus candidates say that nobody else was good enough, partly based on CV and partly based on skype interviews. So I think the size and quality of that 2ary pool can be a big impact on whether a failed search or keep looking decision is made.
I would add that some searches fail because the Department does not really agree what they want. They can agree enough to get a job ad written, but not enough to make an offer. Of course this is also a function of the pool of candidates. No single candidate could satisfy enough of the factions within the department to get an offer. Can happen in many searches, but much more common when there are significant internal divisions to start with.