Does Heat Reduce How Much Work Gets Done?
On hot days output falls and absence rises, but the size of the effect depends almost entirely on whether the work is exposed to the weather.
Reviewed against primary sources on September 8, 2026 by the Soon operations research team. How we vet the evidence
The evidence in one line
In a national panel of Indian factories, annual plant output fell by roughly 2% per degree Celsius, and worker-level data showed productivity declines of 4% to 9% per degree on hot days, alongside higher absenteeism; climate control mitigated the productivity losses (Somanathan et al., 2021). US time-use data found large reductions in labour supply concentrated in industries with high exposure to climate (Graff Zivin & Neidell, 2014). Exposure, not temperature alone, is what determines whether these numbers apply to a given workplace.
The size of the effect where it has been measured best
The strongest single estimate comes from Indian manufacturing, combining a census-scale panel of factories with worker-level microdata from selected firms. Annual plant output fell by about 2% per degree Celsius of higher temperature, and on hot days worker productivity fell by 4% to 9% per degree (Somanathan et al., 2021). Absenteeism rose as well, so the loss came through two channels at once: people present but slower, and people not present.
Those two channels responded differently to intervention. Climate control mitigated the productivity losses on the shop floor, which is the intuitive result. It did not remove the absenteeism, which is the more interesting one: cooling the workplace does not cool the journey to it, the home a worker slept in, or the illness burden of a hot week. An operation that has air-conditioned its floor has addressed part of the problem and should not assume it has addressed all of it.
Why the number does not transfer to every workplace
US time-use evidence found large reductions in labour supply from higher temperatures, but concentrated in industries with high exposure to climate, with similar reductions in time spent on outdoor leisure and a reallocation toward indoor activity (Graff Zivin & Neidell, 2014). The net effect on total US employment was modest precisely because most US work is not weather-exposed. The same paper notes effects could be substantially larger where climate-exposed industries dominate and baseline temperatures are already high.
That is the qualifier that gets stripped off when these figures travel. A per-degree productivity elasticity estimated in weaving sheds and garment plants is not a general law of human output, and applying it to a climate-controlled office overstates the effect badly. It is equally a mistake to dismiss the finding in a temperate country, because warehouses, kitchens, forecourts, construction and delivery work are weather-exposed wherever they are.
What this changes about summer rosters
The operational reading is that heat is a staffing problem before it is a comfort problem. If absence rises with temperature and cooling does not fully prevent it, then a heatwave is a coverage event that can be forecast from the weather like any other predictable demand, and covered with the same planning that covers a known peak.
The second reading concerns which hours exposed work is scheduled into. Both studies describe effects that concentrate at the hot end of the day and the hot end of the distribution, so moving exposed tasks away from peak afternoon heat is the intervention the evidence most directly supports. Neither study tested a specific rescheduling policy, so the size of any benefit from doing so is not established.
What this means for your schedule
- Plan for higher absence during hot spells, not just lower output, because climate control was found to mitigate productivity losses without removing absenteeism.
- Apply per-degree productivity figures only to weather-exposed work; the labour-supply effects concentrated in climate-exposed industries and do not describe a conditioned office.
- Treat forecast heat as a coverage event with a known lead time and roster against it, in the same way you roster against any other predictable demand spike.
- Where exposed tasks can move within the day, moving them away from peak afternoon heat is the step this evidence most directly points to, though no study here tested that policy.
The business case
Heat losses arrive through two separate channels, slower work and missing people, and an investment that addresses only the first leaves the second in place.
Because temperature is forecastable days ahead, the exposure is unusually plannable compared with most sources of absence, which makes it a cheap place to start.
Frequently asked questions
- How much does output actually fall per degree?
- In the Indian factory panel, annual plant output fell by about 2% per degree Celsius, while worker-level data showed declines of 4% to 9% per degree on hot days (Somanathan et al., 2021). The gap between those figures reflects different units of measurement, annual plant output against day-to-day worker productivity, not a disagreement.
- Does air conditioning solve it?
- Partly. Climate control mitigated the productivity losses in the plants studied, but absenteeism persisted (Somanathan et al., 2021). Heat reaches workers through their commute, their sleep and their health, none of which a conditioned shop floor changes, so cooling the workplace addresses one channel of the loss rather than both.
- Do these findings apply to office work?
- They should not be assumed to. The labour-supply reductions found in US time-use data were concentrated in industries with high exposure to climate (Graff Zivin & Neidell, 2014), and the per-degree productivity estimates come from manufacturing. Applying a weaving-shed elasticity to a conditioned office overstates the effect substantially.
- Is there an optimal office temperature?
- Not one established by the studies cited here. These estimate how output and labour supply respond to outdoor temperature variation; they do not identify a setpoint. Figures circulating as an ideal office temperature generally come from separate laboratory work on thermal comfort, and should be checked against that literature rather than attributed to this one.
- Is this the same as saying climate change will cut productivity?
- It is the evidence base such projections are built on, but a projection adds assumptions about adaptation, air conditioning uptake, and industrial composition that the underlying studies do not settle. Both papers found evidence of adaptation, including shifting the timing of activity, which is precisely what makes simple extrapolation unreliable.
Sources
Every figure on this page is drawn from a cited primary source and checked against the original publication.
Somanathan, E., Somanathan, R., Sudarshan, A., & Tewari, M. (2021). The Impact of Temperature on Productivity and Labor Supply: Evidence from Indian Manufacturing. Journal of Political Economy, 129(6), 1797โ1827. https://doi.org/10.1086/713733
Design: Panel study of a national census of Indian factories plus worker-level microdata from selected firms
Graff Zivin, J., & Neidell, M. (2014). Temperature and the Allocation of Time: Implications for Climate Change. Journal of Labor Economics, 32(1), 1โ26. https://doi.org/10.1086/671766
Design: US time-use panel study exploiting daily temperature variation within counties
Cite these sources: BibTeX RIS
Why this page is graded moderate evidence
A consistent systematic review or meta-analysis at a lower grade, or a large observational study whose authors disclaim causality.
Who reviewed this
Every article in this library is checked against its primary sources by the Soon operations research team: each figure is traced back to the study it came from, and the wording is checked against the study design before publication. What that review covers
None of the studies cited here evaluated Soon.They examine scheduling practices, shift patterns, and working hours as studied by independent researchers, so their findings describe what those practices are associated with, not what any particular software produces.
This article summarizes published research for scheduling and operations decisions. It is not medical advice. Individual health questions belong with a qualified clinician.
Keep reading
Does an extra hour of work produce an extra hour of output?
Two panel studies, one from 1916 munitions plants and one from a modern call center, disagree about whether there is a threshold at all.
Read →Does Overtime Cause Injuries, or Do Hazardous Jobs Use More of It?
The largest US cohort found long hours associated with more injuries even after adjusting for industry and occupation, while the meta-analysis that followed is far less settled.
Read →How Do You Judge a Scheduling Study?
A short field guide to reading scheduling research, worked through this library's own sources, including the claims we checked and cut.
Read →Does better scheduling actually pay?
One randomized experiment, two large surveys, and an honest account of how much of this field we could not verify.
Read →Coverage map
Related resource
Open →Forecast tool
Related resource
Open →