Two call centres in China record what each worker produces each day, and Tom Chang matched those records to daily pollution and weather readings. Higher air pollution lowered daily output, and it did so at concentrations that occur routinely in large cities in developed and developing countries alike. The abstract available here gives the direction and the range but no coefficient and no worker count, so neither the size of the effect nor the sample can be stated.
Pawel Wargocki turned the classroom literature into a dose-response curve. The paper does not say how many published studies went into it, and learning was measured through psychological tests, mathematical and language tasks, rating schemes, end-of-year grades and exam scores, and with short-term sick leave included because it feeds into progress. Air quality was represented by carbon dioxide concentration. Reducing it from 2,300 to 900 ppm predicts a 5 per cent improvement on the tests of progress in learning. Reducing it from 4,100 to 1,000 ppm predicts 2.5 per cent higher daily attendance. In ventilation terms that is the range from 2 to 10 litres per second per person. The measurements come from schools, and the curve is the reference an office has to borrow because nobody has drawn one for offices.
Rajat Gupta monitored two offices continuously for 19 months, one naturally ventilated and one mechanically. In the mechanically ventilated office, task scores dropped above 1,000 ppm of carbon dioxide. In the naturally ventilated one the threshold was 1,400, and median scores below it were up to 12 per cent higher. Perceived productivity fell whenever it was too warm or the air was felt to be stale, and arithmetic and proofreading tasks confirmed it. The paper reports the two buildings and the 19 months and not how many people worked in them.
Bluyssen and Sakellaris published two analyses of the same survey. Within the European OFFICAIR project, 26,735 invitations went out and 7,441 office workers in 167 modern buildings across eight countries answered, in the winter of 2011 and 2012. Twenty-three per cent rated their overall comfort below four on a seven-point scale, and among those the main complaints were noise other than from building systems, air that was too dry and temperature that was too variable. An increase in perceived control over the indoor climate was positively associated with the perceived quality of the environment.
Samin Marzban analysed 1,403 post-occupancy surveys from 14 open-plan offices in Australia, New Zealand and Hong Kong. Satisfaction with visual privacy and with access to the outdoor environment were the two strongest predictors of productivity. Layout and interior design predicted creativity. Organisational factors predicted mental health; the building did not.
Only one study here measures production without asking anyone: Chang’s daily output records, and that paper is behind a paywall. Bluyssen and Sakellaris asked 7,441 people about their own productivity, and Marzban 1,403. Gupta measured two buildings for 19 months. Wargocki pools classroom studies, so the dose-response curve for an office is borrowed from a school.