currently reading gad allon’s post about when to arrive at airports (typing as i go, so it’s not going to be the most organized/concise/well-written). he introduces this idea of the newsvendor problem where a newsvendor has to decide how many newspaper to stock each day w/o knowing exact demand. hitting average means high stockouts or excess inventory, and the proposed solution is choosing a quantile of the demand distribution determined by relative cost of each outcome.

so for this airport arrival, studies have found that security wait times are let’s say 20 minutes. to calculate the cumulative distribution of waiting time b/w arrival to gate (F_s), we do C_o / (C_o + C_u) with Co being cost of overage or cost per minute of waiting too long and with Cu being one-time penealty of being late. we then input into F_s L*, which is the optimal arrival buffer or the quantile of the distribution that balances those costs.

thus, averages are misleading as you’d likely be in the 90th-99th percentile where the lower range would be for leisure and higher range would be time-sensitive, high-premium travel. you, as a passenger, would likely not be in 50th percentile as underage cost can be ridiculously expensive as we see with my dear friend Hope who had to pay an exorbitant fare to take the next international flight out to make it back home in time for work a month or so ago. as for the overage cost, even if we assume $100,000 average annual income at ~$48/hr, the additional hour, let’s say, is marginal, especially considering time spent in lounges or browsing shops, which may reduce the perceived cost of overage.

rational travelers optimize to higher quantiles, not the mean, as cost of missing a flight would likely easily dwarf the cost of waiting, especially for international flights.

so, yes, on average, you don’t have to be 2 hours early. but the variance remains too high and the underage cost too daunting that the safest strategy is as the century-old newsvendor says: overstock. whether it be time or newspaper, you’d rather have extra on hand than none at a time of desperation.

for myself, i almost always arrive quite early at international airports abroad. i try to aim for at least an hour between post-security/immigration and boarding, allowing me some time to explore the airport, grab some drinks and maybe food at lounges (w/ priority pass [essentially only valuable abroad]), and prepare all my offline entertainment. for domestic flights, i try to aim to be past security right as we start boarding. there’s no valuable lounge access that i have, and i’m less intrigued by domestic airports. however, aiming for that has led me to arrive at airports like SFO in the middle of boarding, yet I still somehow made it through security & onto my international flights ontime. tbf, they were all cases of poor route time estimation by uber + unplanned traffic incidents + optimistic time buffers; thus, that’s why you aim for the 90th+ percentile, not 50th. anyways, as for airport dad, i’m more convinced he doesn’t want to suffer emotional and mental exhaustion of post-missed-flight drama and chaos rather than the financial damage.

source: gad allon’s article