Journalism grants fund great work. Can we make applying suck a little less?

Preview · Audio edition

Journalism grants fund great work. Can we make applying suck a little less?

0:00

/0:00

1×

You're hearing a short preview. Paying members get the full audio edition of this article.

Upgrade to listen

Wendy Wei is maybe the most grant-savvy reporter I know personally. She uses grant funding to do amazing things. To name just one, she went to Ukraine with support from an International Women's Media Foundation grant and published three stories, on humanitarian aid workers putting their lives on the line and on wartime cultural life, based on her reporting there.

"It really was my dream to be able to bring international stories back to the American Midwest, where I'm from," Wendy told me. But beyond that, she said, "I knew having an IWMF grant would be the stamp of approval I needed to launch my freelance international journalism career." So even though it only covered hard costs — training, travel, insurance, interpreters, housing, transport, etc., and not the cost of her time writing and reporting — she applied. (For the 8 weeks she spent reporting, writing, and handling edits on the stories that grant supported, Wendy was paid only $650 by the magazines that took her stories. One magazine didn't pay at all because it ran out of budget.)

Journalism is, objectively, not doing so hot. Since 2000, the number of people employed by newspapers in the USA dropped by 80%. Freelance rates haven't budged for decades. Cue sad violin music. Because journalism is kind of falling apart at the seams, philanthropic funding — including in the form of grants and fellowships — has become very important for funding the kind of high-effort journalism that everyone says they want but nobody seems willing to pay for when there's free crap available on social media. You know, the Good Stuff. Like what Wendy writes.

Applying to grants and fellowships takes time. At the very least, journalists need to submit a short description of their intended project and — if they're freelance — letters of commitment or support from editors willing to run their stories if they're funded. Often, there's a lot more required than that. Wendy needed 24 hours of dedicated work spread over about a month to put together her application for her IWMF grant to report from Ukraine. Speaking from my own experience, it took over a week of pre-reporting and working on my application, plus a trip to Vienna, to apply to the FRONTIERS science journalism residency. Talking to other journalists, anecdotally, this does not seem unusual. So while grants are great, applying to them can quickly become a bit of a chicken-and-egg problem. You need a grant to report the story, but also need to report the story to get a grant. Tricky.

Though it might sound like grant-ception, one solution to this problem would be to provide grants to fund grant applications. I want to argue for a particular way of doing this that I haven't heard of before.

The closest thing I've seen to what I have in mind is the Fund for Investigative Journalism's seed grants of $1,000-$2,500 for freelancers to support reporting that can lead to full investigative grant proposals. In addition to a resume and two work samples, the application requires a one-paragraph summary of the project, a longer narrative of the potential story, an explanation of why you think there's a story, a detailed breakdown of costs for pre-reporting, and an explanation of other related reporting and how what you're proposing is different. Regarding covering reporting time: "The journalist’s time can be one cost covered by the grant, but it should not be the only cost." (A requirement that strikes me as a bit odd — reporters would be incentivized to add e.g. travel to a story that wouldn't otherwise need it just to meet this requirement.)

However. For a pre-reporting grant to really do its job of making grants more accessible, it has to be a light lift. Applying cannot cost more in time and effort than the grant is worth. So, I propose a scheme that's similar to the FIJ seed grants in spirit, but with a much lighter application (and easier screening process for the grantor) based on a journalist's track record and portfolio rather than their proposed project — and selection by lottery.

The problem: grant math doesn't math

For freelance journalists who want to do ambitious work involving travel or other costly reporting, grants are not just a nice-to-have. If magazines won't cover our costs — which they often won't — then grants are basically the only option. In Wendy's words, they're part of the business model.

I treat applying for grants as a central task of being a freelance journalist, as equally important as the journalist tasks of interviewing, sourcing, fact-checking, etc. Being "freelance" means you're a business owner, and I estimate I dedicate the equivalent of at least a full workday each week scouring for new opportunities, applying, attending in-person meetings/events to keep this little business afloat. 

So let's think about grants from a business perspective. Time to pull out my phone calculator and try not to make dumb math mistakes in a newsletter post that hundreds of people will read:

Let's say a freelancer wants to invoice for $60,000 a year. That's a bit less than the mean salary in the USA (and remember, this is before the freelancer deducts business expenses — self-employed people have to foot their own bills for literally everything from tax accounting to travel to office coffee, printer paper, etc.).

At 40 hrs/week, 50 weeks a year (2 weeks of vacation, no sick days — getting sick as a freelancer is not recommended), we're looking at an hourly rate of $30/hr. Except, not really. Because less than half of freelancers' work is typically "billable hours" — i.e., work you get paid for. The rest is admin and other tasks needed to keep the business running that nobody pays you for. Like looking for grants. If our freelancer is really efficient and 50% of their hours are billable hours, we need to double that hourly rate to $60/hr.

(Note: in my experience, this is a kind of unrealistically low needed hourly rate for freelancers to feel they're doing okay in 2026. But let's stick with it as a conservative estimate. You could also argue that applying to grants and fellowships falls into the 50% of non-billable-hours I used to inflate the hourly rate, but I find it more helpful to compare the cost of applying to a grant to the pay for a billable project since applying to a grant is an income-generating activity like writing an article, not an admin activity like handling your taxes. When I think about whether any income-generating project is "worth it," I calculate the effective hourly rate and compare it to the hourly rate I need my billable hours to hit.)

Let's say our theoretical freelancer wants to apply to the European Geosciences Union's journalism grant, which awards up to €5,000 for travel costs to report on Earth science. The winners this last year got €920 ($1,030), €1,350 ($1,510), and €3,000 ($3,360). Applying requires a 2-page proposal with a fleshed out motivation, reporting plan, and detailed budget and feasibility assessment, as well as a 1-page summary of experience detailing a journalist's professional achievements.

Let's say applying to this doesn't take a week. It takes 3 days of billable hours. So that's $1,440 for our freelancer to apply. The first grant already doesn't pay off even with a 100% win-rate. The second grant effectively pays $70 once you deduct the cost of 24 hours of lost billable work... with a 100% win rate. Drop the win rate to 40% — which would be pretty good odds  — and none of the most recent winners' grants pay off. The full €5,000 breaks even for a success rate of about 26%. And remember, this is for a grant that covers hard costs and none of the cost of the journalist's time. So from a business perspective, things actually look worse.

What about bigger grants and fellowships? It really depends on the odds of success — which are often not transparent. But let's consider a fellowship that does share some information about its success rate.

The Knight Science Journalism fellowship is one of the best fellowships a science journalist can hope to win. It is hosted at MIT and has an MIT-esque acceptance rate around 2.75% (11 winners from a pool of 400). So, even though the fellowship awards a fat stipend of $87,500, the slim odds mean the expected value of applying is $2,406. That roughly breaks even with the $2400 cost of 40 billable hours for our theoretical $60/hr freelancer. And investing just 40 hours in pre-reporting and applying to a program like KSJ is unrealistic. Sure, some journalists might bang out an application in a few days. But the amount of unpaid pre-reporting that goes into gathering the information that makes a 2-day application possible, spread over months and years, is almost certainly more than 40 hours.

But Elise, one might reasonably object. Thinking of this in terms of expected value is stupidly oversimplified. The KSJ — if you get it — is a life-changing, huge thing. Applying is high-risk, high reward.

Yes, you're right. And this is also a problem. A lot of freelancers cannot afford to "spend" a week of work — and lose out on a week's pay — to buy a lottery ticket, no matter how big the jackpot is. Even in cases where the expected value math works out, plenty of journalists simply cannot front the cost of applying. Assuming risk is the privilege of those with means they can afford to lose. So risk ends up acting like a screen. It excludes the people who'd benefit the most from grant funding or fellowships before they even have a chance to apply.

If the subtext isn't clear already, I want to say this explicitly: grant pre-reporting is an equity issue.

An aside on making the math math

People do manage to do great things with grants, and there are ways to make grant funding work as part of a freelance journalist's business model. So in the interest of not painting an unreasonably bleak picture, a counterpoint:

Wendy writes for outlets that often can't afford to pay her much (Aside: as a science journalist writing for magazines that pay upwards of $1.50/word, I cannot believe how little she's paid for work that matters way more than my "ooooh fun physics!!" stories. But that's another issue.). So she funds a lot of her work with grants. It can work. Kind of. She told me that she's at a point where she can mostly re-use her grant applications when seeking funding for particular projects. "So it's not a sunk cost for me," she said. The key is finding something you know you really want to write about and applying to grants again and again with variations on that theme. If 1 week of pre-reporting feeds into 6 grant applications, each application costs less and the math maths better.

Advice I hear often is that you can finance pre-reporting by writing related freelance stories that come with a project fee. This happens naturally when journalists have a developed beat, or a topic that they cover intensely, over and over.

"If you find yourself chasing random grants out in left field, that’s ineffective," Wendy says.

Even if costs can be lowered with clever re-use of material, pre-reporting still presents a problematic barrier to seeking grant funding that's worth lowering.

The actual solution: fix the information environment's financial incentives

Ideally, the financial incentives of the information environment wouldn't be so screwed up and newsrooms would have money to pay journalists to do ambitious important journalism. The problem of grant applications wasting a bunch of money and time would mostly evaporate if grants were a nice bonus rather than a prerequisite for doing big things. People could just get to work, rather than working for free to convince someone to give them money so they can work.

But "fix the financial incentives of the information environment" is kind of beyond my power as a freelancer who writes a newsletter with, like, 400 readers. (Sign up to support my work for less than $1 per post so I can accumulate more power readers.)

So how about a semi-solution that won't fix everything, but might actually be possible to implement?

A feasible semi-solution: a portfolio-based micro-grant lottery for pre-reporting

I propose a program that awards small (~$500-$3,000?) grants to support pre-reporting for more ambitious grants. These grants should cover both hard costs and, for freelancers, pay a stipend for several days to a week of reporting and grant-writing to cover lost income during the time spent applying.

Grants would be awarded in a lottery after a pre-screening process to eliminate unqualified applicants. Being "qualified" would mean 1) being a journalist with a track record of publishing in established media outlets on a topic adjacent to the proposed topic of pre-reporting and 2) committing to applying to a specific journalism grant that requires a fleshed-out proposal.

The point is to make it very quick to apply (hopefully no more than an hour) and very light to screen. For our theoretical $60/hr freelancer, a pre-reporting grant of $1,440 to cover three billable days of pre-reporting for that EGU grant makes applying a good deal every time, even if the odds of winning are low. And if it only takes an hour to apply, the expected value of the $1,440 pre-reporting grant is positive so long as the acceptance rate stays above 4.2%.

More time spent reporting, less on begging for money to be allowed to report. More money and volunteer hours devoted to supporting reporters, less on administering a grant and screening applicants.

That's it. That's the idea.

Aren't lotteries unfair? What about diversity and equity goals?

I mean, define fair. (No, really, it's hard. There are entire research groups dedicated to figuring out and implementing fairness in algorithms for selection processes like grant applications.)

The question we should be asking is not whether lotteries are perfect, but rather whether they're a better tool for the job of awarding micro-grants for pre-reporting than the other available options.

From Wendy: "I do want to say to any naysayers about the random lottery system: there’s so much luck built in to all grants."

Lotteries make sense when you have far more qualified applicants than you have slots for. This is, I suspect, true for most journalism grants and fellowships. It's also true in many cases for science grant funding, which is a reason science funders have been experimenting with lotteries, for instance in Australia, New Zealand, and Switzerland. A lottery doesn't have to be totally open; it can be the final layer on top of a preliminary screening process that sorts out unqualified applicants.

Regarding equity, reducing barriers to entry already helps. A low-lift application is going to attract more diverse applicants than something that requires a week of unpaid work up-front. But beyond that, there's no reason a funder couldn't engineer a lottery to support their equity and diversity goals — or other goals, like priority subject areas. Special pools for applicants that meet certain criteria, like being early-career or belonging to a certain demographic group, are one option.

Why would funders give up control like this?

Because funders who care about supporting quality journalism should have an incentive to use their money effectively. Lotteries with a portfolio as a barrier to entry are a way to do that.

People and organizations with a track record of doing excellent work are unlikely to stop doing excellent work if you give them money. Rather the opposite. Most people doing great things on the regular just want opportunities to do great things. Journalists are in it for the love of the game — we're certainly not in it for the pay.

"There's a lot of talented, capable, dependable freelance journalists out there. We shouldn't waste their time when their track record demonstrates that well enough," said Wendy.

It's understandable that grant funders would be wary about giving up control — they want the biggest bang for their buck, and media nonprofits all want a piece of the philanthropy pie. But allow me to make a counterintuitive suggestion. Perhaps trusting people who do great things to keep on doing great things if you give them money is a more cost-effective, high-impact way to distribute funds than wasting resources trying to hand-pick your way to maximum impact? What if the cost of occasionally funding a dud project (which happens even with rigorous screening, anyway) was less than the cost — to journalists, the volunteers who often screen grants, and grantors — of labor-intensive application processes?

An anecdote: I once worked in a big research group where there was a cabinet full of office supplies that anyone could take for their lab work. But people, of course, took supplies for personal use: notebooks, pens, etc. For a while, there was a tracking system in place and someone on admin staff was responsible for making sure nobody was stealing notepads and extension cables and pens. But it turned out that it was more wasteful to have that person policing the supply cabinet than it was to accept some low baseline rate of thievery. Giving up control was more cost effective.

The cost-benefit analysis will of course vary by program. But as a final point: "lottery" doesn't mean "free-for-all." Funders could set the bar for entry wherever they want it. They could require applicants to have a byline in a publication from a certain list of flashy magazines, have previously won a grant from a flashy funder or a particular award, or even have a personal invitation from the grantor if you want to get really elite and hand-shakey about it.

I am actually trying to make lottery micro-grants for grant applications a real thing that exists in the world — help me? (Or steal the idea?)

I do love complaining about things on this newsletter. Complaining can, very occasionally, make real things happen in the world. But that's kind of the exception, not the rule.

This time, I do actually want to try to do something instead of just complain. Because I feel like a pilot run of a micro-grant lottery — unlike a healthy information environment in the year 2026 — is actually a sort of feasible thing to turn from an idea into a real-life thing.

I've taken my very first baby steps towards putting together a budget request for the National Association of Science Writers (NASW) to fund the pilot as a program for its members. Why? Because this is the path that opened up for me when I shared my idea with a supportive NASW member. To be clear: I have never administered a grant, and I have no clue what I'm doing. I welcome people with better ideas to shred me in the comments — or, if you're feeling more constructive than that, shoot me an email to help shape the budget request for the grant pilot.

Ultimately, though, I don't much care who makes this happen. I just want it to happen. So if you hold the purse strings of a nonprofit that funds journalism, or sit on the board of a journalism professional society, or maybe even are a scientist who wants to use journalists as low-cost guinea pigs for a study on how lottery-based grants work or whatever — be my guest. Seriously, please do it. You can even say it was your idea, I do not care. Ideas are cheap. Money and management and logistics — and time — are expensive.

Minor Revisions

Updates on developing stories, corrections, and things I missed

The AI slopocalypse is overwhelming academic publishers and destroying the journal article as career "currency" for science. Last newsletter, I wrote about how the arXiv preprint server put a 2-per-month hard cap on new submissions in response to a deluge of AI-generated submissions. It is, of course, not just the arXiv. More just from last week: The American Astronomical Society just announced that its journals are rejecting planet discovery papers that follow an LLM-generated formula after being inundated with submissions over the last two weeks. (More in plain languge in astronomer Chris Lintott's Bluesky thread). Anna Rogers, a computer science professor, shared on LinkedIn that her main takeaway from the EMNLP conference on natural language processing last week was "exhaustion and depression" due to reviewing being under-incentivized and authorship (now eased with AI) being over-incentivized. Anna Wexler, an associate editor of the journal AJOB Neuroscience, wrote an opinion piece for The Atlantic about how the journal article as an academic "currency" is collapsing in a flood of AI-generated submissions. Related: Jan Pisek and Eve Veromann argued that hallucinated citations in the academic literature should be treated as academic misconduct in a a correspondence to Nature.

The Trump admin's attacks on US science makes the NYT. The New York Times ran a huge story with 7 journalists on the byline about the Trump administration's catastrophic science policy. By huge, I mean it takes 25 minutes to listen to the audio edition. If you've been reading this newsletter (or, at least the Citation Needed section) for a while, you'll know there's a lot to talk about. Related: Dan Garisto, Max Kozlov and Mariana Lenharo wrote a piece for Nature looking ahead at what the upcoming US midterms could mean for science.

AI math: OpenAI and Anthropic in a proof "duel" and more. The two AI companies produced proofs refuting the exponential time hypothesis — a foundational idea in computer science with practical implications for cryptography and a step towards the famous "P=NP" problem. P=NP is a Millenium Prize problem, like the Navier-Stokes problem reportedly solved by an OpenAI model last month. There's a lot more to say about AI math that I missed during my scattered travel weeks, including progress on the Riemann hypothesis that was among the 700-ish results OpenAI "carpet bombed" the mathematics community with last month. (Perhaps in response to criticisms frmo mathematicians, OpenAI had several mathematicians tidy up the exponential time hypothesis to make it more comprehensible to humans.) AI math results aren't just a problem for math, like how to handle authorship on scientific papers reporting results mostly or entirely generated by AI models. See also: an arXiv preprint posted last week that set frontier GPT models on a random sample of 200 open mathematical conjectures and found that the models produced candidate solutions 60% of the time.

Science media sad violin noises. The Wall Street Journal is cutting about a dozen staff from its health, science, and education desks. If I haven't said it enough times already, the financial incentives of the current information environment does not reward good journalism. Science is often seen as a luxury beat to cut when things get tight.

Citation Needed

Last week in links, quotes, and commentary. Inclusion is not endorsement.

• So... about grant lotteries. Is it time the ERC used one? Applications to the European Research Council's elite science grants swelled by 60% even as funding constraints cut the number of available grants. Researchers are starting to feel like winning is a lottery. The expected success rate is now less than 5%. (With so many qualified applicants, an actual lottery might actually be fairer and less burdensome.)
• Internal documents reveal Meta mulled over meddling in science on social media harms. Documents revealed in a lawsuit show that the social media giant was playing with the idea of intentionally starting a whisper campaign against an outspoken critic, Jonathan Haidt. Haidt is not a great dude. But that doesn't somehow excuse Meta trying to discredit an academic because his research and writings threaten their business — if it can happen to him, it can happen to others. The documents also show Meta employees apparently considering whether the media or academics would be easier to manipulate, and seeming to settle on the media. Academics like Joan Donovan have spoken out before about Meta running interference on their research on social media.
• Record-breaking retraction: An engineering society pulls 1,400 papers in one go. The Institute of Electrical and Electronics Engineers (IEEE) announced last week that it had retracted 1,417 conference papers previously linked to paper mills by a team of academic sleuths. The papers were identified as paper mill products by their titles —which had been advertised as papers for sale on social media.
• What's going on with a major open-access journal? Retraction Watch reported that the journal Scientific Reports stopped being indexed in MEDLINE, a major biomedical library. Why? Not clear yet. It's easy to speculate that it has something to do with the ballooning number of papers published in Scientific Reports, which is open-access and charges a $2850.00 article processing fee.
• Lessons from the UK's metascience unit. Jen Gold shared a few takeaways from her experience as co-head of the UK Research and Innovation (UKRI)'s metascience unit, which runs studies on how to run science effectively. One funny takeaway: appealing to researchers' "sense of social obligation" in emails inviting them to review grants did not increase their willingness to volunteer.
• Europe is putting €16 million into a preprint-first publishing model. Holly Else reports for Nature on an investment EU countries are making in Open Research Europe, a "publish, review, curate" model for science that makes publishing easy, supports open peer review, and allows users to easily curate the research they want to see. It's a pretty radical approach that doesn't play nice with a lot of academic measuring sticks like the journal "impact factor."
• Science is an attention economy. Exhibit A: Dinosaurs. Fossils of dinosaurs, mammals, and other "glamorous" creatures are over-represented in top journals, Ewen Callaway reports for Nature.
• Scientists in the West aren't as optimistic about their jobs as in India and China. That's one of several interesting results on how scientists feel about their careers from a recent Nature survey.

Making Progress

A study published in Science reported notable structural similarities between birdsong and human speech — and that's very useful to actually know, because birds are used as a model organism for studying human language.

Get in Touch

Send me your tips, comments, questions, concerns, and definitely noteworthy theories of everything: hello@reviewertoo.com

Or just drop a comment below