Broadband's waiting-time payoff concentrates among middle-income Americans

Work backward from the two headline numbers and a baseline appears. If a 1.19% cut in waiting time is worth 1.5 hours a year, the average person in the sample was waiting roughly 126 hours a year, about 2.4 hours a week ^[1]. That is the quantity broadband is being asked to move, and a 1.19% slice of it is small enough that the estimate's precision matters as much as its direction. The Conversation labels the piece a Research Brief, a short take on academic work ^[17], and it reports neither standard errors nor the time-use data behind the waiting measure ^[16].

Start with the denominator under the national figure. Divide 36.5 million hours by 1.5 hours a person and you get about 24.3 million people, which is the 24 million Americans the authors say lack broadband ^[2][5]. The valuation itself is arithmetically clean: 36.5 million hours at $33.54 comes to $1.22 billion ^[6]. Per person, the same inputs give 1.5 hours and about $50 a year ^[4][3], priced at a national mean wage rather than at what the households in question earn ^[7].

Reconciling the women's numbers is harder. Solve (11 + x)/2 = 1.19, assuming women are about half the affected population with similar baseline waits, and everyone else has to be waiting roughly 8.6% longer ^[5], which the brief does not report. The hours sit more easily: 14.4 million out of 36.5 million is 39% ^[4], close to what a female share of the population would produce if women's per-person saving were near the overall 1.5 hours rather than the 9.2 times it implied by the coefficients as printed ^[7]. Most likely the two figures come from different specifications, and the brief does not say which.

County broadband adoption is not randomly assigned, and the authors write in the language of correspondence and association rather than cause ^[1][10]. Their mechanism is specific: agencies, clinics and businesses put more services online, and the queues thin for the people who still show up in person ^[3]. What that design cannot separate is a county where adoption is high from a county where services were better resourced to begin with. The study's own subgroup result is the useful check on it. If connectivity were doing the work evenly, low-income residents would gain too, and they do not; their waits stay long whatever their county's adoption rate ^[11]. The authors say they did not test why, and name subscription costs and digital skills as suspects ^[12].

That cuts against the obvious reading of the result. Time poverty falls hardest on lower-income women ^[8], who have 20% less leisure time than men and wait on other people's behalf as well as their own ^[9]. The measured saving, meanwhile, is largest for middle-income people ^[11]. The authors call the gap the second digital divide, a difference in the ability to use access rather than in access itself, and argue that infrastructure money needs literacy and affordability programs attached to it ^[13][14]. I would treat the direction of the average association as real, and the 11% figure for women as a number to keep out of any business case until its specification is on the page.