“I didn’t get much done last week because it was so hot.”
It wasn’t the first time a client said this to me, and I was curious. “Have you considered getting an air conditioner if it’s that bad?”
“No” He replied (let’s call him Philip), “An open window is usually enough. It’s just that the heatwave this week was particularly bad.” When we discussed it a bit more, Philip said it didn’t seem worth the hassle since the AC only really felt necessary for a few weeks during the summer.
So, was Philip right? Is it worth buying an AC as a productivity hack?
How much is your time worth?
Most people don’t think about air conditioners when evaluating their productivity, because it’s not within the normal scope of time management. But if you frame productivity as the careful use of your resources - including time, effort, and money - to accomplish your goals, then an AC is fair game as a productivity hack. The question becomes “is the cost of buying the AC a good deal for the productive time it saves you?”
The same logic applies to any purchase that increases your productivity. So, to start answering that question, let’s consider the value of an hour. If you don’t already have an estimate of much money you will spend to save an hour, try estimating it with the tools below. A dollar value is a handy tool when deciding which time savers are a bargain.
This Clearer Thinking tool is convenient for calculating the value of a marginal hour of your time. (A normal work hour may be worth more or less to you, depending on whether you care more about output from focused work hours or work/life balance. The marginal output of an additional hour worked decreases, probably starting somewhere after three to six hours per day. On the other hand, the marginal value of an hour of free time to you starts increasing the more you’ve worked.)
Another way to calculate the value of a marginal hour of work time is to look at what you could earn in a marginal hour as a baseline for how much your time is worth, e.g. tutoring on Wyzant. These estimate how much you would be willing to pay for an hour of your time. If your income is particularly unstable or you have less than 6 months runway, you may want to be more cautious about spending money.
You could also try directly estimating the value you produce in an hour. Since this seems difficult to do for most people, a plausible proxy is how much existing metrics value the impact of your time. 80,000 Hours’ survey found EA org leaders valued their most recent junior hire for 3 years at $1,050,000 (average) and $450,000 (median). Assuming 40 hours a week, 50 weeks a year, that values an hour of junior hire time at $175 and $75, respectively. For a senior hire, they gave $7,400,000 (average) and $3,000,000 (median), which works out to $1,233 and $500 an hour respectively. While these numbers may be overestimating the value, an order of magnitude less would still be $17.5 per hour for a junior person and $120 per hour for a senior person.
What does an AC cost?
Now that you have an estimate of your time, let’s compare the value of an AC to other ways you can buy productive time with money. For these calculations, I’m going to estimate an hour is worth $30 (50 cents a minute). Your personal value for time may be much higher (or lower if you’re cash constrained right now).
Research on the impact of heat on productivity suggests productivity starts decreasing at as low as 22°C (72°F) and decreases by about 2% for every degree Celsius above 24°C (75°F) (Footnote 1). This would imply that at 29°C, it would take about 4 extra hours per week on average to get the same amount done (a 10% reduction in productivity). While exactly where your productivity starts decreasing depends on your gender, it seems almost certain you’re not doing your top work if it’s blazing hot in your office.
Let’s do some quick back of the envelope calculations. Assuming that your productivity is reduced by 20% per week when the temperature is above 35°C (losing 8 hours per week) and you value your time at $30 an hour, that heat caused you to lose $240 worth in one week.
A window or portable AC unit costs around $300 on Amazon (plus up to $20 per week in electricity) (Footnote 2). That buys time at about 35 cents per minute in two weeks of a heatwave. If you have four weeks of heatwaves, your cost per minute saved goes down to 20 cents. After that, the cost of running the AC for an entire week is more than paid for by saving 1 hour.
If you really don’t want to buy an AC, you could accomplish the same productivity benefit for hassle instead of money. E.g. find a place that does have AC and work there during heat waves. You could also try working in the morning and evening, and taking a break during the warmest parts of the day.
Additionally, I expect a lot of the benefit of an AC to come at extreme temperatures, which is where most people report a clearly noticeable difference in productivity. This particularly true if you live in Europe, which is currently experiencing its third heatwave this summer. July’s heatwave broke at least four countries’ heat records with temperatures up to 108 °F.
If you live in a city where the temperature is above 35°C (95°F) for more than two weeks per year and you don’t work in an air-conditioned space, you should probably buy an AC. For more moderate high temperatures, say 26°C (80°F), it seems more likely you’ll be able to cool off with an open window or a fan.
Still, 20 cents per minute seems like a pretty good deal on time, given we started out willing to pay 50 cents. Let’s check some other ways you can buy time.
How much does it cost to buy time other ways?
Paying for a laundry service
A laundry service that picks up your clothes costs $30-$40 a week based on a quick look at two sites, and I’m ballparking that it would save about 30 minutes per week (it’s hard to tell, since it’s in 5-10 minute chunks here and there). So using the laundry service buys time at around $1.17 a minute. Pretty expensive compared to our baseline of 50 cents.
Buying Soylent
Replacing breakfast power smoothies with Soylent. Soylent is $3.25 per bottle ($2.53 if you buy in bulk with Amazon subscribe and save). It takes about 5-10 min and $2 to make a power smoothie with a similar number of calories. So Soylent buys time at about 10-20 cents a minute. So basically, you’re getting an awesome deal on your time by drinking Soylent occasionally. It’s like the Black Friday sale on time. (Note: There are other health considerations that may easily trump the productivity gain, plus taste and variety. I’m also not confident enough in Soylent to recommend drinking it exclusively.)
Eating takeout
I’m estimating $15 to buy a meal and assuming leftovers to be about another meal’s worth for myself. So $7.50 per meal for compared to $2 per meal and 10 minutes (based on 60min spent preparing 6 meals in bulk). So at about $5.50 to save 10min, eating takeout buys time at about 55 cents per minute, almost breaking even.
Hiring a house cleaner
Our house cleaner is $25 an hour, and I estimate she cleans as quickly as I could. So this buys time at about 42 cents per minute, which does a bit better than breaking even. (Avoiding particularly unpleasant hours may be worth more to you than a normal hour, however.)
Taking an Uber
The value of an Uber can vary a lot between rides, but basically it’s worthwhile when you can save two minutes for each extra dollar you spend above what you would have paid for public transit. A recent 30-minute Uber with 3 people cost us $28 ($16 more than the $12 the Bart would have cost), and saved 15 min. Split across 3 people, that works out to 36 cents per minute.
How do those compare?
By these calculations, Philip should drink Soylent (sometimes), buy an AC, hire a housecleaner, and take the Uber.
Of course, the math may work out differently for you. It’s likely that there’s significant individual variation in the value for an hour and the details for the above estimates. Your calculations may look significantly different, so feel free to make a copy of my calculations here and adjust for your own inputs. If you can’t easily measure the input (such as how much heat affects you), try running a personal experiment to test.
Because these specific numbers aren’t the important take away here. Even if none of the above conclusions apply to you, the framework for trading money for time does. So try looking around your life for other opportunities to buy some time at a bargain.
__________________________________________________________________________________
Note: Thanks to Aaron Gertler and Mary Wang for editing on this post! An older version is posted here on lesswrong.com.
(Footnote 1)
https://www.sciencedirect.com/science/article/pii/S2542519618302377
Pg. 527
Workers' health and productivity under occupational heat strain: a systematic review and meta-analysis
“heat stress (defined as wet-bulb globe temperature beyond 22.0 or 24.8°C depending on work intensity”
“11 studies with 8076 workers were included in meta-analysis three. The pooled proportion of individuals showing productivity loss due to occupational heat strain during work in heat stress conditions was 30% (21–39; appendix). In addition to the prevalence of productivity loss, seven studies reported precise changes in productivity as a function of environmental heat stress. These studies suggest an average 2.6% productivity decline (individual study estimates: 0.8%, 1.4%, 1.8%, 2.2%, 2.8%, 4.4%, 5.0%) for every degree increase beyond 24°C WBGT.”
https://www.vox.com/2018/7/27/17611940/heat-wave-2018-cost-workers-deaths-health-climate-change
“Economist R. Jisung Park reported that worker productivity declines by 2 percent for every degree Celsius above room temperature”
“Researchers at the London School of Economics found that urban areas also pay a price for high temperatures, even among indoor office workers. They found that in London, a warm year could cost the city’s economy upward of 2.3 billion euros in productivity. Workers make more mistakes and act more slowly as temperatures rise above their optimal range.”
Summaries of studies on temperature and productivity which found “remarkable consistency in point estimates: (at least) -2% per degree C above room temperature.”
Table 4
This technical report claims hourly work capacity drops by 74-90% when the temperature is 37°C (98.6°F), compared to a baseline of 26°C (78.8°F).
(Footnote 2)
Utility costs, likely overestimating the monthly cost since Jeff’s electricity is 3x the national average.