What Do Americans Do for Fun?
(the code of this project can be found on Github:Β R_code)
The American Time Use Survey:
Working in the marketing industry, I've always had consistent interest in observing peopleβs lifestyle: How do they use media? How do they spend money? And of course, how do they spend their time? With my deep curiosity about this particular topic and passion for data analysis, I discovered ATUS, the American Time Use Survey.
American Time Use Survey (ATUS) is sponsored by the Bureau of Labor Statistics and is conducted by the U.S. Census Bureau. The major purpose of it is to develop nationally representative estimates of how people spend their time. ATUS respondents are selected from the Current Population Survey(CPS) samples, and they are interviewed only one time about how they spent their time on the previous day, where they were, and whom they were with.
Β All about the data:
The original ATUS data comes with 6 linked files:
- The Respondent file (Contains information about the responding individual)
- The Roster file (Contains information about the respondentβs whole family)
- The Activity file (Contains information about the activities respondent reported)
- The Activity Summary file (Contains activity and demographics related data)
- The Who file (Contains information about the people who are with the respondent while the activitiesβ in process)
- The ATUS CPS file (Contains more demographics that comes from CPS data)
I discarded the who file and roster file for better focus, joined the remaining four datasets, and Β kept only the observations from 2011 to 2015. Β Β The variables used in this study include those related to demographics, activity time use, education, job and income.
After data cleaning, the raw dataset came down to 1,118,532 observations of 33 variables.
American time usage in 5 years:
Speaking of Time Use, the first thing I looked at was the trend along a timeline. Β The bar chart below shows the top 10 activities over a five year period from 2011-2015. I initially hypothesized that certain healthy activities, like working out, would show a gain in popularity year after year, but unfortunately, the answer was NO.
In this chart, I summarized and ranked the total duration time of each independent activity and divided it by the unique respondent number of that year. This gives us the top 10 time consuming activities and, approximately, the hours spent on that activity by each American / each day. We can see, of course, the number one activity was sleeping. One interesting finding is that, on average, Americans spent more time watching TV and movies than working! At first glance I could not believe my eyes and went back to check the code. What I found is that the survey blends weekdays and weekends, and does not exclude the unemployed. But anyway, this finding is valid enough to indicate the irreplaceable status TV and movies have in American culture and the impact on peopleβs life.
Thatβs partly why we have the word βcouch potatoβ, which isnβt a favorable description of a person at all. Β To discover more in this realm, I decided to drill down into recreational activities specifically and find out what are the other things Americans do for fun.
So, what do Americans do for fun?
Using the same method, I listed the top 10 recreational activities that Americans spend most time on. Note that a missing bar means that particular activity appears in the top 10 list at least once but is replaced by another activity in that year. The pattern stays consistent and, again, TV & movies were more heavily used than any other leisure activity.
I was surprised to see that attending sports activities, like going to a ballgame, Β barely made it into the top 10. But I was even more surprised to see that no sports activities made it into the top ten as I tend to think of Americans as active sports lovers. Β The data told me that it's just my stereotype and it seems most Americans prefer a sedentary lifestyle.
Next I do a comparison between people with different backgrounds and locations: Region, state, income and education level.
Good leisure and bad leisure:
The first thing I need to do is to set up a standard of healthy lifestyle in terms of recreational activities.
I found that together the top 10 recreational activities possessed nearly all of the time people spend on leisure, so it is fair enough to look at only the top 10s since they are representative enough.
I divided them into 2 categories: Beneficial & Non Beneficial.
Beneficial Recreational Activities are: Social, Reading, Relaxing & Thinking, Doing Sports, Attending or Hosting Social Events, Attending Sports, Arts & Crafts as Hobby, Other Arts & Entertainment.
Non-Beneficial Recreational Activities are: TV & Movies, Shopping, Computer Use for Leisure, and Games.
This is for sure a disputable classification, but there is barely any standard criteria for this, so I made the classifications based on my judgement. Β
Regional rivalry on healthy life style:
Applying the definition above, I calculated a new variable, which is:
the Time spent on Beneficial Recreational Activities / Time spent on all Recreational Activities
I call it Beneficial Fun Time Ratio, and use it to gauge the healthiness level of one's entertainment life.
The graph above clearly shows the relationship between Region, Family Income and Beneficial Fun Time Ratio. Β There are several interesting conclusions we can draw:
- Generally, people living in the west have the most healthy recreational life and southern Americans prefer non beneficial fun choices.
- It is very strange that northeastern residents with low family income have the lowest Beneficial Fun Time Ratio of all when the overall level of northeastern is pretty high.
- $50,000 is a watershed where family income starts to show a possible correlation (needs to be tested, of course) with Beneficial Fun Time Ratio. From this point on, the higher family income, the healthier they are in terms of the way they entertain themselves.
States rivalry on healthy lifestyle:
However, it is simply not enough to look at only the region, what about states? Can I rank the states by Beneficial Fun Time Ratio?
On the map above, darker cyan indicates better performance on Beneficial Fun Time Ratio.
The result is very interesting and sort of counter-intuitive. I was supposing with sunshine and ocean, Californians and Floridians would have a very high ratio. Β On Youtube videos is looks like they surf and bike all day long. Β I even went back to check my code and data manipulation process, but unfortunately, in the world of data visualization, counterintuitive things happen all the time. The only reason I can think of is that maybe the geeks in California spend too much time playing games.
Does education level make a difference?
I also want to examine the possible relationship between lifestyle and education level. Thus I plotted the graph below, and I also added a new continuous variable, personal weekly income as another factor.
The first thing everybody will notice is that people who did not finish college have obviously less income than those who did. Β But that's not the purpose of this graph. In this particular graph, dots gaining density towards the top would indicate that this education level has higher Beneficial Fun Time Rate, but that never happens. It seems that some doctors can possibly be couch potatoes and some people with lower degrees can maintain a healthy lifestyle too! So my conclusion here is Education Level does not have a powerful impact on lifestyle when it comes to recreational activities.
Key takeaways:
- Americans spend a lot of time in their life watching TV & movies
- Location and income definitely shape a person's entertainment life
- Education level does not affect the healthiness of one's recreational choices
- Be careful when you surmise one state must have healthy/ unhealthy lifestyle, since the reality can be totally counter-intuitive.
Next steps:
Given the dataset, I can also explore the relationship between different job titles, industries and the Beneficial Fun Time Rate. I also need to drill down to each state's micro data to explain the map I plotted previously. In all, this is such a fun data exploration journey and if you read till here, I would like to thank you for your patience and you've probably been sitting for too long, now you may need to get up and go for a walk!
About Author
Leave a Comment
