A recent scientific study has revealed the answer to this question for us. I don’t know about you, but I am happier when it is neither hot nor cold. I have lived where it gets cold: Upstate New York and Wisconsin and where it gets hot: South Central Texas and West Tennessee. Living under those extreme conditions is one of the main reasons I fell in love with Western Oregon 23 years ago. Here we don’t see either of those extremes very often and when we do they don’t last very long.
A January 7th article from the Washington Post referenced on Wonkblog written by Christopher Ingraham pointed to the research of Patrick Baylis who is a 5th-year PhD candidate in agriculture and resource economics at the University of California at Berkeley. Baylis is also affiliated with Dr. Solomon Hsiang’s Global Policy Lab and the Electricity Markets and Policy Group at Lawrence Berkeley National Laboratory.
The article summarizes the research by quoting Baylis this way “He found that, compared with a day when the high temperature is 72.5 degrees, a day with a high temperature of 90 degrees makes the typical person experience a drop in happiness similar to the drop in happiness between Sunday and Monday.”
“Temperature and Temperament: Evidence from a billion tweets” is the title of the working paper written by Patrick Baylis and released through Energy Institute at Haas (EI). If you have never heard of them, according to EI, “Working papers are circulated for discussion and comment purposes. They have not yet been peer-reviewed or been subjected to review by any editorial board.”
The question to me becomes how do you get adequate data to get these results? I remember back when I was working on my independent study project to get my degree at Memphis State University, now known as the University of Memphis in Memphis, Tennessee, I performed a survey. Since I had never done this before I contacted the Psychology Department since they were considered the “experts” when it came to the do’s-and-don’t’s of surveys. The result was that I assembled a survey to be done over the telephone. I wanted to personally hand out the surveys but was told that statistically it would not be random enough. So I ended up making over 3,000 telephone calls to get 401 people to answer the 4 questions on my survey. That is what you might call the “old school” way.
Patrick Baylis took a completely different approach. The term social media is relatively new and Baylis decided to use it to his advantage. He used the “Twitter Universe” that was created in 2006. Something I never thought of is that tweets are considered in the public domain so no permission is needed to use the data from them. He used a Twitter client’s Streaming API account to get access to massive numbers of tweets from all over the United States. The tweets were “Geo-located” which means those who posted the tweets gave permission for their location to be revealed. The data collection took place from June 2014 through October 2015 with two breaks in between. That is a good starting point, but a way to decode the tweet needed to be devised.
The analytical model has four parts. The first is called Expert and uses a dictionary (AFFIN-111 dictionary) created by experts that translates the words of the tweet into expressions of positive or negative feelings of happiness in response to the current weather conditions. A total of 2,477 words comprise this dictionary. The result is a measure ranging from -5 to +5 to determine the hedonic state (state of happiness) where -5 is a strongly negatively hedonic state (negative happiness) and +5 is a strongly positive hedonic state (positive happiness).
The second part is the Crowd-sourced measure. It is similar to the Expert model, but the dictionary contains 10,000 words, used by the Mechanical Turk service. It differs from Expert due to its using words whether or not they have a subtle indication of the state of their happiness.
The third part is the Emoticon measure. It uses the emoticons that subjects added to their tweets to give additional insight into the happiness state of the tweeter. A lesser portion (about 2%) of the tweets surveyed used emoticons so a specialized formula was used to combine the emoticons and the happiness-related words to determine the person’s happiness state.
The fourth and final part of this complicated formula is the Profanity measure. Baylis assembled a list of over 300 profanities and gave a score for the presence or absence of profanity in the tweets. Profanity in tweets most often is an indicator of a lower level of happiness.
He then added the weather component to the mix. Data of mainly temperature and precipitation came from the PRISM Climate Group’s AN81D data set. He used data from 2,162 weather stations around the 48 contiguous United States including sky condition, visibility, relative humidity, barometric pressure and wind speed.
I read every word of the more than 50-page paper and I can attest that it is written in the standard academician’s language that most of us would strain to understand upon first reading without looking up many phrases and terms. So, lets sum up the results as simply as possible.
In his discussion section at the end of the paper Patrick Baylis states: “I find that hedonic state is unaffected by cooler temperatures, but declines sharply above 70 degrees F.” So, in his study the participants showed more happiness with the cooler weather and much less happiness as the temperature increased. He also comments that air conditioning is so prevalent in the United States that the unhappiness levels of people in other parts of the world where they don’t have air conditioning as available could be even much greater with the increasing temperatures.
Wow! That’s a lot of work to attempt to prove that most people would rather shiver than sweat. I also noticed that there was no mention of Seasonal Affective Disorder (SAD) which can have a negative effect on happiness during the Winter months particularly here in the Pacific Northwest.
If you have an idea for a future topic let me know what you would like me to talk about or explain. You can comment below or email me at: [email protected].