At a glance
Postdoc talent sourcing has become a global contest shaped by funding shocks, visa costs and rising pay floors. A practical guide for PIs, talent leads and research offices, with a sourcing checklist and where to advertise.
Postdoctoral recruitment was for decades a byproduct of the seminar circuit. A principal investigator heard a promising doctoral candidate present, a conversation followed, and a two-year appointment took shape well before the paperwork existed. That route still works, and in well-networked departments it works smoothly. What has changed is the number of points at which it can now break.
The published statistics suggest a healthy market. The National Center for Science and Engineering Statistics counted 69,877 postdoctoral appointees at US academic institutions in autumn 2024, a 6.1 per cent rise on the 65,850 reported a year earlier and the highest figure in a survey that has run since 1966. The previous peak was 66,247 in 2019. Health sciences grew fastest at 9.6 per cent, ahead of engineering at 5.5 per cent and the sciences at 4.6 per cent, with biology and the biomedical sciences alone accounting for 20,234 appointments and clinical medicine for 17,919.
Headcounts, however, describe decisions taken years earlier, when grants were awarded and cohorts admitted. They say very little about the market a research team is sourcing in this month.
What the 2025 numbers do not show
The National Postdoctoral Association polled its members in May 2025, after the first round of federal terminations. Roughly two per cent of its 23,000 members responded, and of those, 197 reported that their federal funding had been cut, up fifteen points on an earlier poll taken in the administration's first month. The share reporting that they had lost their positions rose from 4 per cent to 11 per cent between the two surveys. Twenty-one per cent said visa or immigration processing had obstructed them.
A survey of 378 people cannot characterise a workforce of seventy thousand, and the association has been careful not to claim that it does. It does not follow that the sample is uninformative. Self-selected respondents tend to be the people with something to report, which is precisely why the direction of travel matters more than the decimal place.
What those numbers describe is a market being rearranged by forces outside any department's control, and a hiring committee that plans as though the 2019 market still obtains will lose candidates it never knew it had.
The pool has not vanished, but the terms have changed
In a nationally representative survey of academic scientists conducted by researchers at Arizona State University, 70 per cent reported a reduced ability to recruit international students or postdocs since January 2025. Sixty-three per cent said they had warned team members about international travel risks, and 55 per cent said they had edited keywords in research proposals. The SciOPS survey findings describe a hiring environment in which the constraint sits upstream of the interview.
A separate poll of 989 researchers across 45 states, conducted for STAT News by the MassINC Polling Group, found that 14 per cent had seen candidates turn down offers because of immigration policy, and 13 per cent had lost researchers to institutions in other countries following funding cuts. Those are small proportions of a large workforce, and they translate into a great many unfilled posts.
The machinery has also become more expensive. A presidential proclamation signed in September 2025 attached a $100,000 fee to new H-1B petitions, up from a few thousand dollars, and was upheld in federal court that December. A second proclamation, issued in December and effective from 1 January 2026, suspended or partly suspended visa issuance for nationals of 39 countries. Most postdocs enter the United States on J-1 exchange visas rather than H-1B, so the fee does not capture the whole picture — but departments that have historically converted a postdoc to an H-1B, or recruited a candidate already on one, are now budgeting for a number that changes the case for the post entirely. George Washington University's guidance to departments is blunt about the timing: about a month for a domestic hire without a visa, and three to six months, sometimes longer, where sponsorship is required. The GW postdoc hiring guidance is unusually explicit about this and worth reading as a model of what candidates need told.
The result is not a closed border but a reordering of preference. Growth in the number of international research scholars at US universities slowed to 3.5 per cent last year from 10 per cent the year before, according to NAFSA, the association of international educators. A QS analysis of the shift quotes a higher education policy specialist warning of gradual erosion rather than sudden collapse. An analysis by Nature found applications from US-based researchers to posts abroad up 32 per cent in the first quarter of 2025 against the same period a year earlier.
Photo by Brecht Corbeel on Unsplash
The case that there is no shortage deserves a hearing
Principal investigators reading the foregoing may reasonably object that their inboxes do not look like a crisis. In a survey of 34 chemists, biologists and physicists published in EMBO Reports, respondents reported receiving between 20 and 70 applications for each advertised post, many of them excellent. Choosing between them, the author noted, was far from trivial. What those PIs said they wanted was motivation, evidence of independence, and technical expertise that matched the project. What put them off was a generic salutation, inflated claims, and a cover letter that could have been sent anywhere.
It does not follow, however, that applicant volume equals candidate reach. A post that draws sixty applications from the same five doctoral programmes, all of them already known to the committee, has not been competitively filled in any meaningful sense; it has been filled from a network, which is a different procedure with a different set of risks. The GW guidance names this directly: a PI who has identified a candidate whose training closely matches a specific programme may appoint without a competitive search, provided a written justification is filed. That is a legitimate instrument, and it is also one that quietly narrows the pool a field draws from.
The shortage, where it exists, is of fit rather than of applicants. A post requiring experience with a particular imaging modality, a named cohort dataset, or a language spoken in the field site will not be filled by sixty generic applications. Neither will a post that depends on a visa that takes five months to arrive and whose grant has eighteen months left to run.
Pay floors are being reset while federal scales stay flat
Compensation is now the first filter candidates apply, and they apply it across borders with a currency calculator open. The US federal benchmark, the Ruth L. Kirschstein National Research Service Award scale published by the National Institutes of Health, runs from $62,232 at zero years of postdoctoral experience to $75,564 at seven or more for awards made under the 2025 notice. The NIH stipend notice is the document most institutions quietly reference when setting their own minimum; Penn State, for instance, sets a floor of $61,008 for postdocs appointed from autumn 2025.
Collective bargaining has pushed several campuses well above that floor. The University of Washington's postdoc contract set a minimum of $68,460 from January 2024. Columbia's postdoctoral union reported minimums of $72,116 for postdoctoral research scientists under the contract that has since expired, against a union demand for a floor of $80,000. A candidate comparing two adverts, one of which publishes a number and one of which says "competitive", will assume the worst about the second, and reasonably so.
The competition is not only American. Canada announced a C$1.7 billion research talent initiative in December, intended to recruit up to 1,000 researchers over twelve years with multi-year funding attached. European universities have been advertising in English on the same portals as their US counterparts and, in several cases, moving deadlines forward to catch candidates whose plans changed. Germany illustrates the risk in recruiting aggressively without retaining: around 90 per cent of postdocs under 40 there hold temporary contracts, and one analysis puts the exit rate at 46 per cent after the first postdoc year, rising to 66 per cent after five. A survey of 872 postdocs across 74 Max Planck institutes found rates of depression and anxiety higher than in a comparable survey three years earlier, with international researchers — roughly three-quarters of the cohort — reporting the sharpest difficulties.
For a research office, the implication is unglamorous. Retention data and recruitment data are the same data, and the institution that cannot say how many postdocs it employs, on what contracts, funded from where, cannot answer the first question a good candidate asks.
A sourcing checklist for the current cycle
Most of what follows costs nothing but a decision. None of it requires a new scheme or a policy change above departmental level, which is why it is worth doing before the next competition round.
Photo by Brecht Corbeel on Unsplash
- Write the methodological requirement as a sentence a stranger could verify. "Comfortable running multilevel models in R on cohort data of 10,000 or more" recruits better than "strong quantitative skills", and it lets a candidate in an adjacent discipline recognise herself in the advert.
- Publish the salary or stipend figure, in the currency of payment. State whether the appointment is a stipend or an employment contract, because in most systems the two carry different benefit consequences that candidates discover only after accepting.
- Name the funding source and the end date, then say what happens at the end. "Two years, renewable subject to funding" tells an applicant the second year is a coin toss. If a bridge scheme exists, name it and state its conditions.
- Put the visa position in the advert. Whether sponsorship is available, which category, and whether the appointment is contingent on it. Silence is read as refusal, and international candidates read it first.
- Run rolling review rather than a fixed deadline when the post is grant-funded. Tell shortlisted candidates within three weeks. A four-month committee process is a subsidy to better-organised competitors.
- Search two pools at once: the people who read the board you posted on, and the corresponding authors of the last five papers in your specific niche. Send the advert to those authors directly and ask them to forward it.
- Ask your own postdocs who they would hire. They know the cohort one step behind them, and they have nothing invested in the network you already have.
- Keep the file on everyone you shortlisted and did not appoint. A grant that lands in eight months will need it, and the candidate you could not fund this year is the fastest hire you will make next year.
Where a global search actually reaches
The checklist above widens a search only if the advert travels. AcademicJobs.com carries postdoctoral, research, faculty and administrative listings from institutions across North America, Europe, Australia, New Zealand and Asia, and routes applicants back to the institution's own site rather than holding them inside a portal. Its postdoc jobs section and research jobs listings are the two places a grant-funded appointment should appear, and postings can be live quickly enough to catch a candidate who is deciding this month rather than next quarter.
The platform's matching tools scan candidate profiles when a post is published and notify those who fit, which matters in fields where the strongest candidates are not browsing boards daily and are instead being courted by two other institutions. Employers receive a shortlist rather than a slush pile. Research offices handling several searches at once can use the recruitment tools to keep the pipeline visible across departments.
The practical move is to treat the advert as a document with a job to do. If your last postdoc search ran through three personal emails and a conference corridor, run the next one through both that corridor and an open listing, then count where the applications came from. Post the post with the salary band, the funding source and the visa position stated, and see what the wider pool returns before the grant clock forces a decision.
The terms of the advert are the precedent
Every postdoc advertisement is a small statement of policy. It says who is welcome, how long the welcome lasts, what happens when the grant ends, and whether the person appointed will be treated as staff or as a line in a budget. Multiplied across a few thousand departments, those statements determine which researchers can move between countries this year, and therefore which questions get worked on.
The record number reported for 2024 was produced by decisions taken in 2019 and 2020, when awards were made and cohorts admitted. Whatever is decided now, in shortened contracts, withdrawn fellowships and searches that quietly go to the candidate already in the building, will show up in the statistics around 2030, by which point the people who left will have built careers elsewhere. The precedent worth watching is not whether the postdoc survives as a stage of training. It is whether it survives as an international one.
- academic-jobs
- higher-education-recruitment
- international-researchers
- principal-investigators
- postdoc-talent-sourcing

