Pursuit of Happiness “What can data tell us about happiness”.
Happy countries are all alike; every unhappy country is unhappy in its own way.
In a world where growth often equates economy, OCED raised the importance of well-being, which put country-level happiness in the spotlight.
Today I’m looking into 2016 World Happiness Report commissioned by UN, which contains global happiness index and a series of social-economic factors such as economy, family, health, freedom, trust and generosity. Countries are increasing adopting Happiness as a measure of social progress. It’s widely aware that Bhutan uses Gross National Happiness (GNH) as a development indicator. To less public awareness, Mar 20 has been designated as International Day of Happiness.
First thing I noticed is a broad pattern where countries rank higher on each factors tend be happier. So after some data cleaning, I looked into how countries in each regions are doing in terms of their social-economic well-being.
Some observations:
- generosity, trust and freedom all tends to be rather low
- some region are more homogenous (such as North America) than others (such as South America).
- there are outliers in the region: trust index is rather low in Southeast Asia, except Singapore; economy index in Eastern Africa is especially high in Mauritius, which is above world median too; Sri Lanka and Bhutan are much better off in family index compared to their South Asian peers.
This is the first step to visualize the dataset. Coming next I intend to look into how social-economic well beings contribute to country-level happiness and how does the trend change year-on-year.
What I learnt today is that in order to loop through charts, one needs to wrapplot in print() statement. The charts are simply bar-charts projects onto polar coordinates.
You could read more about world happiness report here.
I’m looking into happiness disparities among regions.
According to the survey, world’s happiest country in 2016 was Denmark and the unhappiest country was Burundi.
Happiness by region
When we go up the scale of happiness, countries in these regions stand out: North America, Western Europe, ANZ.
In search of the happiness countries
And we see happier countries tend to be better off on social-economic metrics too. The measures themselves made me curious, as they almost seems to be individual-focused, as we don’t see country-level metrics like political stability.
Distribution of happiness score
Happy to an different extent
We observed:
- North and South America appear to be most happy on average, followed by Europe, Asia, Africa and Middle East
- Africa and Middle East has a high outlier: Isreal
- Latin America and Carri beans has a low outlier: Haiti
Relation between happiness and its drivers
To understand how individual drivers relate to happiness score, we model linear trend line of each country and label some of the outliers by region
Economy vs Happiness
Family vs Happiness
Freedom vs Happiness
Life expectancy vs Happiness
Trust vs Happiness
Generosity vs Happiness
We can observe corruption and generosity have weaker correlation with happiness compared to other factors.
Clustering countries by social economic factors
Next we can do k-means clustering on countries’ social economic factors.
the countries formed 6 clusters
Cluster 1 contain Middle East, Eastern Europe and Africa, cluster 2 are Latin America, Southeast Asia and Eastern Europe, cluster 3 contain Western Europe and the affluent part of Middle East and Asia, cluster 4 contains moreLatin American and Caribbean as well as Central Europe, cluster 5 contains more European countries and cluster 6 are mostly African countries.
Full code on my github.
Thanks for reading and welcome to send me ideas and suggestions.
