World Happiness Findings

Trends, analysis, explorations and further research proposal for the most recent World Happiness Report which is generated by United Nations Sustainable Development Solutions Network.

All data cleaning, exploration and analysis was done by me using python 3 in a Jupyter Notebook and all data visualizations were done using Matplotlib, a visualization library in python.

Introduction

If we could boil down the 7.5+ billion people down to few key tendencies that most contribute to happiness, what would they look like? What would this tell us about who we are as a collective? What would it tell us about what we most value and what, ultimately, correlates most with increased feelings of happiness within a country’s population? I had the impulse to want to choose certain countries for this project, but I thought that looking at all the countries represented in my data at once could lead us to some interesting and intuitive conclusions about what most of us seek, in varying phases and degrees, throughout our life journey: happiness. 

Correlation

Further research will need to be conducted to try and find causality, but since we’re at the beginning of our discovery we can begin with a few basics. We’re looking at 159 countries across the world with metrics in the following categories: 

• Ladder: Measure of life satisfaction.
• Positive affect: Measure of positive emotion.
• Negative affect: Measure of negative emotion.
• Social support: The extent to which Social support contributed to the calculation of the Happiness Score.
• Freedom: The extent to which Freedom contributed to the calculation of the Happiness Score.
• Corruption: The extent to which Perception of Corruption contributes to Happiness Score.
• Generosity: The extent to which Generosity contributed to the calculation of the Happiness Score.
• Log of GDP per capita: The extent to which GDP contributes to the calculation of the Happiness Score.
• Healthy life expectancy: The extent to which Life expectancy contributed to the calculation of the Happiness Score.

Taking a birds eye view of this data all at once is a bit daunting, but if you take a quick glance below, you see that there are few concentrated areas of strong correlation between some of the above variables. The four most notable pairs include: "Log of GDP per capita" & "Healthy life expectancy", Social support & "Log of GDP per capita", "Healthy life expectancy" & Social support, and “Freedom” and “Positive Affect” were the closest clustered pairings with the clearest, most linear relationship contributing to the factors that bring people greatest happiness. 

Health, Happiness & Prosperity

How many times have we seen the classic “Health, Happiness & Prosperity!” on a greeting card? When we want to express sincere well wishes, the best we can come up with is a wealth of health, happiness and joy (presumably spent with those nearest and dearest) and prosperity and a feeling of abundance, of having just a bit more than you need for your wants. What about freedom? Indeed without one of these, we’re off. But why is that exactly? Taking a look at a few of these greater insights can show us a bit about our relationship to happiness and its’ intrinsic components. 

Key Tendencies 

Coming back to the four most notable pairs mentioned above: "Log of GDP per capita" & "Healthy life expectancy", Social support & "Log of GDP per capita", "Healthy life expectancy" & Social support, and “Freedom” and “Positive Affect”, we see a strong connection between these sets more so than any other. From the following four groupings we can make the following observations of factors that contribute most to happiness:

1. GDP correlates strongly with healthy life expectancy, along with social support 
2. Social support is proportional in countries with healthy life expectancy and GDP 
3. It appears freer societies have an increased positive affect

More Questions!

These tendencies leave us asking more questions. Which comes first, GDP growth or high life expectancy? Are people living longer and thus contributing further to GDP or are they living longer because the economy is healthier and is improving quality of life? Are countries with greater social support more successful, or does the success of the nation create more social cohesion? Do people feel more positive because of freedom in society or do negative societies breed more authoritarianism? Does life expectancy grow due to increased social bonds or do we create more social bonds because we live longer? 

Heat Map

Perhaps taking another look at how the tendencies relate to each other could be best served with a heat map. Below, we see areas of high correlation and areas of low correlation: 

Among the lowest areas of impact we will find Corruption. Could this be because given the other areas of measuring quality of life and happiness in a country, corruption has the lowest felt impact contributing to happiness? Among the warmest areas of the map, we yet again see close relationships between social support, GDP and healthy life expectancy as we observed in the previous graphs. A great area to explore hereafter would be to look at why corruption has such a weak impact on average for most countries. Is corruption expected and ignored? Or perhaps is it so out of sight and behind the scenes that it doesn’t impact people?

Top & Bottom 5 Analysis 

We can use analysis of the top 5 (Finland, Denmark, Norway, Iceland and the Netherlands) and bottom 5 countries (Rwanda, Tanzania, Afghanistan, Central African Republic, South Sudan) in our list of 156 countries to examine a few telling trends from our list. We’ve chosen the following categories to distinguish factors that contribute to the happiness of these two groups: Positive affect, social support, negative affect, healthy life expectancy, 

Curiously, some of these categories give us some surprising results. The bottom five countries derive happiness most from Generosity, Social support, Positive affect and Healthy life expectancy as compared with the top five countries. What does this tell us about these societies? Perhaps that they are societies built with interdependence as a shared value which in turn impacts how they view their overall quality of life (skews more positive) as well as the way they view their health and life expectancy. Of the categories we’ve examined, the bottom five countries score higher in negative affect which might not be as surprising considering this had a higher impact on their happiness than the top five. 

Further Research Proposal

In order to answer some of the questions posed throughout this report, we will need to uncover additional data that could help us explore some of these relationships and try and distinguish causality from correlation. Uncovering which aspects are drivers can tell us a lot about some of the social benefits of strengthening our economies and prioritizing health and well being the way New Zealand recently announced it would. 

Our comparison of the top and bottom five countries in our list also gives us some additional points to want to follow up on. The most central of those points being: what can the most vulnerable countries teach us about how human beings connect and rely on one another. Why are their contributors to happiness so much more different than the top five countries? What can they teach us about the shared experience of living in a country with high insecurity. 

Could it be true that the citizens of those countries prioritize each other more and don’t take the blessings they do have for granted. Could hardship in turn create more happiness due to a greater sense of purpose? Could the top countries tendencies show us that comfort and opportunity breed further individualism, loss of social dependence and complacency? 

Research Proposal Outline

The Problem: 

There is confusion over which of the aspects of world happiness which we looked at are causing certain relationships or if they have a more complex relationship. Understanding more about which of the studied metrics contribute more to the happiness of the countries included in our dataset would help us with perpetuating happiness for citizens in these countries in the future.

We know that happier citizens in countries would do much to boost wellbeing and even productivity in ways that are innumerable, qualitative and often behind the scenes. While this is a nebulous undertaking, if we can get closer to understanding these relationships, we can get closer to quantifying some of the dynamics previously unseen. 

The potential solution: 

One potential way we can gauge causality is to administer a survey to those that originally provided answers to the UNDSC explicitly asking for causality so that we may gather enough data on the contributing factors to happiness specifically. Without more exploration into the driving factors, those most directly linked to an increase in happiness, we wouldn’t be able to determine causality so this could be a natural next step to determining causality.

The method of testing the solution:

Rather than understanding these relationships through the context of which factors are most active to their happiness, we can ask the surveyed population to rate the top five contributors to their happiness. This would allow us to quantify which factors specifically would produce an increase in happiness for those surveyed. Assigning a descending value of 5 to the top contributor, 4 to the next highest and so forth would allow us to give each factor of happiness a quantifiable value that we can use for further research. 

The first step in the design would be to craft the top five qualities. Given the key tendencies mentioned above, we can use the values in each set to try and gauge hierarchy. We would be asking those surveyed to rate the following factors in a scale from 1-5, 5 being the top factor: "Log of GDP per capita", "Healthy life expectancy", “Social support”, “Freedom” and “Positive Affect”. To account for any inherent bias we are unaware of, we could administer the survey twice at six months apart. If there is more than a 5% difference, we could average the delta. 

Suspending disbelief and exaggerating the realm of possibility, we could also go as far as to make large scale changes to these factors themselves within the countries we’ve seen indexed in the world happiness report 2019 and see if the participants exhibit any changes to their answers. Because we are working in a very subjective capacity, having iterations and steps would allow us to derive insights from any changes in recorded data.

The participants in the first part of the experiment would likely not rate their answers accurately. In order to account for this, we would need to create a scenario where there would be a control group as the observed data could be too subjective. In order to simulate a randomized, controlled trial we could identify all the factors we want to rate, in our case the five mentioned above, and analyze the observed data from the initial survey in order to tease out how important one factor is relative to the others. 

From this second set of data that we would gather after there have been substantial changes, we could turn our attention to two pairs of countries that could have similar dynamics. In our case, we have selected two countries from the top five and bottom five list: Finland and Norway, as well as Rwanda and South Sudan. Our interest in selecting these two pairs would be to analyze their similarities and differences in efforts of finding some composite tendencies between the two pairs.