Programs that start at 18 are screening for something high school already decided

The null result carries most of the analytical weight here. Adolescent friends who went on to build formal, incorporated companies predicted later entry into both incorporated and unincorporated self-employment ^[6]. Friends who simply worked for themselves as solo contractors or freelancers predicted nothing, not even entry into the same informal work they were doing ^[7]. That rules out the easy version of the story, in which any self-employed adult in a teenager's field of view makes working for yourself thinkable. Something narrower is moving through the network. The authors' candidate is visibility: when a former classmate registers a company, hires staff or raises capital, the news reaches old friends through a reunion, a mutual acquaintance or a social media post, while freelance work stays small and relatively invisible ^[11][13]. And people weigh information from someone they have a real connection to more heavily than the same information from a stranger ^[12].

Set that against the intake form. Futurpreneur's core startup program, like most of the field, opens at 18 and expects the applicant to show up with a business idea already in hand ^[3]. Both requirements encode an assumption that the entrepreneurial impulse arrives in adulthood ^[15]. If the association Kwon and colleagues report is real, an idea at 18 is not a starting condition. It is an output of the previous decade, and the idea requirement is a proxy screen for whoever already had a classmate worth watching. Capacity added to a program built that way moves more people through the same filter. It does not widen the pool the filter draws from.

Which is why the demand number is easy to misread. Applications rose 50 percent last quarter against 15 percent growth a year earlier, more than three times the earlier rate ^[1][16]. But that surge is happening under a high youth unemployment rate, with more young Canadians considering self-employment ^[2]. Push out of the labour market and newly seeded intent produce the same line on the applications chart, and nothing on the form distinguishes them.

The earnings finding is the part that should unsettle program metrics. People whose teenage friends built incorporated companies earned more from their own ventures in midlife, even after accounting for whether they had incorporated themselves, and the gap held among people running the same type of business ^[8][9]. So two founders with comparable ventures diverged on earnings according to what their old classmates had gone on to build. What travels is the scale of the ambition, not only the decision to start. A program that reports businesses launched is counting the cheaper half of the effect, and by the authors' own prior work, formal growth-oriented entrepreneurship is where the larger economic returns sit ^[10].

Two caveats the source is candid about. These are associations across two cohorts born roughly four decades apart, one nationally representative sample followed into participants' late 30s and one Wisconsin cohort tracked from 1957 to age 65 ^[4][5]. Replication across generations that entered very different economies is worth something. But the article flags a finding on friendships that persisted into adulthood as complicating a straightforward social-influence reading, and the text available to us breaks off before explaining it ^[14]. The dating of the mechanism is the contribution. The curriculum is not in evidence.