A Two-Day Workshop That Was Still Working Fourteen Years

Over twenty-five years ago, Athena set out to build something simple: a community where women in STEM find each other, learn from one other, and rise together. A quarter century later, the room is still full of the best and brightest our field has to offer — and every so often, it's worth asking the question we ask of everything else in business: does this actually work?

Not the surveys, where successful women say networking helped them — those measure memory, not outcomes. We mean the studies with a control group. There are fewer of them than there should be. What they show however, is more specific, and more demanding, than most career advice admits.

In 2004, the American Economic Association launched a two-day workshop for untenured women economists. Demand outstripped seats, so organizers randomized: some applicants got in, some didn't. Fourteen years later, researchers tracked down where everyone had landed.

The women who attended were 11 percentage points more likely to hold a tenure-stream job than those turned away — and 15.9 points more likely to hold one at a top-100 department. They were 12 points less likely to land at an unranked institution. From two days.

Look closer, though, and the story gets stranger. Whether a woman got tenure at all barely moved. Where she got it did. The workshop didn't lift women over a bar — it moved them up a tier.

The easy explanation is contacts: put untenured women in a room with senior economists, and they'll meet people. That's what most of us mean by networking.

The easy explanation is wrong. A follow-up study traced the mechanism and found attendees did gain roughly three additional pre-tenure co-authors. But across the entire sample, only twenty papers came out of a workshop-born collaboration — and once researchers controlled for co-authorship, the effect on top-journal publishing didn't move at all. The introductions were real. They weren't what did the work.

What's left is information: how to pitch a paper, which journals will quietly cost you a year, what a first offer is actually made of, and what happens when you ask for more. The unwritten rules travel freely among people who already have them — and reach everyone else late, or never. Closing exactly that gap is what Athena has spent 25 years doing.

Compare that to the advice women hear most: find a sponsor, someone senior who advocates for you in rooms you're not in.

One study isolated what sponsorship does on its own — a lab experiment: 322 people, one hour, arithmetic problems, sponsors who were anonymous strangers identified by a number. The authors call it a proof of concept and ask for field trials nobody has run yet.

Within those limits, the result stings. Sponsorship raised men's willingness to compete — concentrated among the men worst at the task. It did not raise women's. It did raise women's estimate of their own ability, by about as much as it raised men's. The vote of confidence arrived. It just didn't convert into anything.

Encouragement and information are not the same commodity. Only one of them moves a career.

One is encouraging and narrow: a randomized trial on a U.S. online training platform sorted more than 2,700 unemployed women into peer groups of six — some all-women, some mixed. The all-women groups finished on time more often, certified more often, and found work in their field far more often: 24% against 9%. The population is entry-level and the setting is remote, so don't stretch it further than it goes — but it's the strongest field experiment anyone has run on who's in the group, and the group made of women won.

The other is a warning: in 2015, France required at least 40% of each gender on academic hiring committees. A study of 455 committees at three French universities found women's hiring odds got worse, not better — and the entire effect sat in committees chaired by men. The author stops short of calling it backlash, and notes something awkward: in the fields most affected, committees had already been favoring women before the law arrived. What the quota removed may have been a thumb already on the scale. Either way, the lesson holds: an intervention built to help can do the opposite. The only way to find out is to measure.

Three cautions, house style. First, the economics workshop moved women between institution tiers, not the overall odds of tenure — two of its results sit at the 10% threshold, and one of those fails a robustness check. Second, the sponsorship study is a one-hour lab task, and for two of its three treatment arms, the male-female difference isn't itself statistically significant — a weaker claim than it first appears. Third, the training-cohort trial ran 2,305 women in the all-women arm against 444 in the mixed one, measured outcomes by self-report, and its authors say it applies best to groups with no shared task — not what most of you work in. The direction across these studies is consistent. The precision is not there yet.

Here's what we can't tell you: nobody has run a controlled trial on a networking event, or at least not one that follows anybody's career progression. The research on conferences measures who asks questions, not who gets hired/promoted. So when we invite you to an industry wine reception, we can't point at a number and promise you anything.

What we can point to is the shape of the thing that worked, twenty-five years running: not visibility, not encouragement, but specific, usable, slightly inside information, moving between women who happened to be in the same room, no transaction attached. That's the whole premise Athena was built on. It's why this community keeps showing up — and why, twenty-five years in, it keeps getting sharper.