The data set is sourced primarily from the 2020 World Happiness Report’s manipulated Gallup World Poll (GWP) data from 2005-2019, where the years are represented across variables. The primary variable of interest is the “happiness” score of each country, calculated by averaging responses from 2017-2019 to the following GWP question based on the Cantril Scale, an evaluative assessment of well-being: “Please imagine a ladder with steps numbered from zero at the bottom to 10 at the top. The top of the ladder represents the best possible life for you and the bottom of the ladder represents the worst possible life for you. On which step of the ladder would you say you personally feel you stand at this time?” Then, the effect of six key variables in explaining happiness is determined using regression and a model (equation) is formed, capable of predicting a country’s average happiness score given its scores for each of the six variables, explained below. Our critique of this method is three-fold:
We believe that happiness is temporal, and well-being can’t be defined universally. We feel that the ontologies used within the WHR should be carried through to the title of the report instead of creating a catchy tagline that is open to wide misinterpretation in media coverage.
Our skepticism pushes us to explore the definition of “happiness” within the World Happiness Report and to conduct further research on the confounding variables that affect the dataset. Beginning with the problematic nature of measuring happiness under abstractionist ontology, we found literature and reports that cover “well-being” and “happiness” as two distinct phenomena using the principles of Set Point Theory and Positive Psychology. There are gaps in well-being coverage, even those extending into empirical studies, “indicate an insensitivity to place and context, a determined humanism” (Smith and Reid, 2018, 816). In Re-Thinking the Cultural Codes of New Media: The Question Concerning Ontology, researcher Srinivasan coins the term “fluid ontologies” to help determine categories based on factors that are important to the communities rather than defined by the researchers.
Our project presents a critical interpretation of the data provided by the 2020 World Happiness Report, which is supported in part by the United Nations’ Sustainable Development Solutions Network. We seek to uncover the data silences that stem from the WHR’s effort to quantify, define, and explain “happiness” as a monolith across national, cultural, and geographic lines with a limited ontology.
One’s definition of happiness is undoubtedly influenced by their cultural and societal perceptions and what they hope to accomplish by quantifying it. Before attempting to measure it, we must define it — and that is the hardest part.
Within research, happiness is often (but certainly not always) studied under the umbrella term “subjective wellbeing” (SWB), defined as the “subjective enjoyment of one’s life as a whole” (Veenhoven, 2017).
SWB is often (again, not always) viewed as the link between the following components of happiness, sometimes imagined as opposite ends of a SWB spectrum (Stone and Mackie, 2013):
How a person feels on a day- to-day or momentary basis.
Also known as experienced or hedonic well-being, it is often split up into “positive” and “negative” affect when researched (Veenhoven, 2017).
The extent to which a person feels they’re satisfied with their life.
Also known as a life evaluation, this component of happiness is concerned with long-term, overall satisfaction with life (Veenhoven, 2017).
Whether SWB should be an acceptable measure of a concept as weighted as happiness depends on who you are talking to. Despite its widespread use, SWB as a term faces criticism for attempting to bridge starkly different models of positive experience without a standard structure (Khalil, 2018).
Indeed, as hinted above, happiness defined in terms of the subjective wellbeing spectrum, unfortunately, doesn’t appear to be strict enough to measure for analysis. Throughout our research, we regularly came across literature on happiness that addresses only one of the above components and not the other, depending on its goals — including the World Happiness Report.
Though the Gallup World Poll collects data for both affective and subjective components, the WHR only utilizes the latter (through a question based on the Cantril Scale) to make its claims about happiness (Gallup, Inc. 2020).
The impact of this will be detailed under our first research question, but resonates throughout the project, asking: can there even be a universal quantification of happiness across national, cultural, and individual identities?
Existing ontologies dominate the discussions, measurements, and explanations of happiness — disproportionally representing certain types of (people’s) happiness more than others.
Positive affect is a subcategory of the affective component of subjective wellbeing. It is defined by the 2019 report as the averaged binary responses to the following Gallup World Poll questions addressing laughter and joy:
This momentary, affective treatment of experience stands in contrast with the life ladder score, which utilizes the popular subjective life evaluation assessment, the Cantril Scale, to quantify long-term wellbeing (as defined in our data critique).
Both measures have advantages and disadvantages — some severe in their potential to silence and skew. Most notably, the Cantril Scale has been found to correlate strongly with income, while affective measures are more associated with time spent socially (Gallup, Inc., 2020).
This interactive scatter plot visualizes the relationship between positive affect and ladder score using WHR data collected for the 2019 report. The countries are colored by region, and regional averages for both scores can be visualized by clicking any corresponding data point. The lines indicating the average positive affect and average ladder score form quadrants, allowing us to see that scoring above average for one of the indicators does not imply scoring above average for the other.
The immediate impact of tweaking our definition is striking. Under positive affect, 2019’s “happiest” country in the world would have been Mexico, and the top eight spots would have been occupied exclusively by countries in Latin America. Over thirty countries would have scored above Finland — who leads the ladder score rankings with other Western European countries — when ranked by laughter and enjoyment.
The studied explanations for the shift in rankings will be considered in the second research question, but for now, the importance of ontology should be made clear.
It should be noted that the World Happiness Report’s use of ladder score over positive affect may be linked to their overarching goals. Its focus on quantifying sources of long-term well-being instead of moments of joy makes sense when considering its relationship to the UN’s Sustainable Development Solutions Network.
The way the WHR wields the word “happiness,” however, feels irresponsible. By ranking happiness through ladder score, only one side of the story of happiness is told — the one that is correlated with high income and healthy life expectancy — allowing existing silences to be exacerbated.
Further research on confounding variables that affect this dataset led us to uncover gaps in well-being coverage, which “indicate an insensitivity to place and context, a determined humanism” (Smith and Reid, 2018, 816). A community-based solution to the issue of definition may be in order, as argued by researcher Srinivasan in Re-Thinking the Cultural Codes of New Media: The Question Concerning Ontology, in which he coins the term “fluid ontologies” to help determine categories based on factors that are important to the communities, rather than defined by the researchers.
“Core concepts from within the eudaimonic and hedonic traditions, examined above, such as the daimon or eudaimonic ‘true self’ and self-reported measures of SWB, betray a similar ontological assumption of an independent preexisting subject, to some degree existing in isolation from spatial context. According to leading scholars in this tradition, wellbeing is an individual or subjective state, ‘a person’s cognitive and affective evaluations of his or her life’ (Diener et al., 2004: 63), encapsulated in terms of cognitive states (such as satisfaction with one’s marriage, work and life) and ongoing affect. Population-level analyses, in turn, comprise an aggregation of these individual scores (Izquierdo and Mathews, 2010: 5; Eckersley, 2008)” (Smith and Reid 2018, 815).
The WHR claims that when taken together, the six variables explain about 3/4ths of the total global happiness. The following bar chart breaks up each of the variables’ explanatory power by region.
This faceted bar graph tracks the average percentage of a country’s ladder score that could be statistically “explained” by each of the six variables, averaged by regional indicator.
Though the WHR warns that the variable explanations for happiness are tentative, the graph above reveals interesting regional information about whose happiness correlates with certain variables more than others, sometimes differently than expected. For example, the Generosity variable predicts more happiness in South Asia, Southeast Asia, and Sub-Saharan Africa than other regions, despite the privileged, arguably Western interpretation of Generosity measured by the question, which asks about “donations made to charity.” We question whether more happiness could have been explained had the question been formed using a less material-dependent definition (Helliwell et. al 2020, 17) .
The average happiness explained per variable is also telling, and is acknowledged by the WHR, which explains global happiness through the following process. First, the WHR creates a fictional country called Dystopia, whose values are the lowest recorded national averages for each of the six key factors, to use for comparison. Every country’s ladder score is higher than Dystopia — which has a 1.79 — and that “extra” happiness must have a reason. The WHR found that the global average of extra happiness is 3.5 ladder score points, which can be explained by the six variables as such (Helliwell et. al 2020, 18):
We ask (and attempt to answer, in the third question): how would different variables (and variable definitions) have changed the conversation around well-being?
Through the Cantril Scale’s correlation with income and the Generosity variable’s assumption of wealth when measuring kindness, the ontologies used by the WHR skew towards interpreting well-being — and therefore happiness, under this definition — in terms of material. By ranking happiness by ladder score and not affect, many experiences of happiness — mostly those in non-Western countries — are silenced.
It’s important to mention that, though the World Happiness Report continues to (somewhat misleadingly) use the term happiness, many scholars have shifted away from it in their research because of its ephemeral quality (Khalil, 2018). By using it, the WHR risks public misunderstanding of its goals — which are likely related to its connection to the United Nations Sustainable Development Solutions Network (SDSN).
The WHR reveals its ontology’s abstractionist quality in the way it defines happiness and the methods it uses. The ontology of “[a]bstractionism assumes that all things, including the self, are the most real and the best understood when they are abstracted or separated from the situations in which they occur” (Slife and Richardson 2008, 701), placing less importance on context. The authors go on to argue that as a result of this, abstracted perceptions of principles or truths place an unwarranted emphasis on similar qualities. Such an emphasis creates the perception that these principles and truths are to be considered universal or unchangeable across cultures, overlooking their important differences (Slife and Richardson 2008, 702). Similarly, the WHR seems to take its definition of happiness as absolute, applying it across the various contexts that exist within each nation. They choose to do this rather than take the relational approach, which more strongly favors context when assessing a concept’s real or fundamental status. However, this ontological abstraction obscures the experience of an individual’s lived emotions. The norms that guide the report lead them to utilize an abstractionist ontology, and in doing so, they limit the ways in which nuances of self and experience are communicated to those who read the report.
By comparing countries’ ladder score to the happiness predicted by the WHR’s regression model, we were able to discover how differences between individualist and collectivist societies reveal the impact of how culture influences one’s perception of happiness.
Residual = Actual Happiness - Predicted Happiness.
In other words, the residual is the difference between the country’s measured happiness and how happy it’s “supposed to be” based on how it scored for the six factors and the regression model (Helliwell et al. 2020, 18).
This symbol map depicts the relationship between a country’s residual score and its geographic location. Interpreting this map is two-fold: 1. countries with higher happiness than predicted have positive residuals (indicated in pink, darkening when the happiness score is particularly higher than predicted) and 2. countries with lower happiness than predicted have negative residuals (indicated on the map in green, darkening when the happiness score is particularly lower than predicted). The size of the dot indicates the happiness score.
This bar graph averages the residuals by region, quantifying the relationship between residual and geography visible on the map. Do note that differences within regions, such as in Sub-Saharan Africa, are hidden on this graph but clearly represented on the map. We can see now that while Latin American countries are underpredicted, Asian countries of all subcategories are overpredicted.
The regional differences between the residuals is striking — and reveals how much “happiness” has not been explained equally by the WHR’s predictive model. From the map, we see that almost all Asian and West African nations’ happiness is overpredicted, and how much of Western Europe, ANZ, East Africa and the Americas’ happiness is underpredicted — particularly Latin America.
Despite generally ranking lower on typical well-being indicators such as inequality, crime, and corruption, Latin American countries lead the world in positive affect and rank “surprisingly” (per the residuals) high in ladder score as well.
Rigorous analysis in the 2018 World Happiness Report points to a cultural tendency towards strong relationships and social foundations for explanation. Daily connection with a tight-knit extended family and friends is apparently the norm, as well as the encouragement of warm interpersonal relationships.
The 2018 WHR also noted that the lack of correlation between positive affect — which is particularly important to Latin American happiness — and ladder score shows that the latter is not a complete measure of happiness in the region, and encouraged further investigation.
Do note that Latin American countries also rank high in “development” and per capita income, other typical well-being indicators. The WHR acknowledges that the region is not perfect — that political corruption and crime weighs down both positive affect and ladder score in the region. Nothing occurs in a vacuum, but we highlight the importance of community, as it has been shown to be particularly influential both in Latin America and globally (Helliwell et al. 2018, 115-121).
On the other side of the globe, many countries’ ladder scores — particularly in Asia — are surprisingly overpredicted, many by over one point. Our goal is to fill in that gap through literature. What additional factors are missing from the existing six that aren’t accounted for in the model, but would better predict some countries’ happiness?
Isolating the data by regions, we explored the importance of culture in our understanding of how happiness is evaluated by individuals. Through our research, we examined a trend toward homogeneity and the landscapes in which hegemonic culture shapes and is shaped by communities.
In individualistic societies, more typical in the Western world, well-being is thought to be derived from self-fulfillment and actualization. In collectivist-oriented societies, common among Asian countries and in Latin America, individuals tend to be attuned to the needs of their communities. Some studies show that under some measures, this may make a difference — individualistic societies like the United States often predict lower well-being, while collectivistic societies like Russia and East Asia predict higher well-being (Snyder, Lopez, and Pedrotti 2011, 27-33).
Despite that, many scholars have argued that not all communities are conducive to happiness for everyone — especially if one lives in a society that operates differently than their beliefs. Some studies have shown that holding individualistic ideals may lead to less happiness in nations that identify more as collectivist — such as in Japan — where some outsiders reported fewer friends and lower subjective well-being (Ogihara and Uchida, 2014).
The commonality between our research, interestingly, seems to be the lack of universality in the measuring and reporting of happiness — contrary to the universality that the World Happiness Report represents. Happiness has been shown to be very much contingent upon socio-cultural ideals at both a national and individual level within a country’s culture.
In this section, we discuss additional factors impacting happiness through visualizations and through supporting research.
Studies have shown that there is a level of social comparison that factors into an individual’s subjective happiness. In one interesting case, people said they would rather have half the purchasing power as long as it meant they were better off than those around them (Solnick and Hemenway, 1998).
Consider the Relative Income Hypothesis, which says that happiness becomes more sensitive to relative income (within one’s comparison group) rather than absolute income once basic indicators of well-being in terms of clothing, food, and housing are met (Khalil, 2018). This phenomenon — that pay comparisons factor directly into predictions of pay satisfaction — has been found across nations, strengthening the idea that at a certain point, one’s numerical income may not impact happiness unless it varies significantly from peers’. (Sweeney and Mcfarlin, 2004).
To quantify the above on a national level, we used the World Happiness Report’s Gini from Household Income data collected by the Gallup World Poll, as shown below.
This scatter plot + trend line illustrates the relationship between inequality within a nation and its average well-being. Though the R-squared for this linear model is about ~0.135, it reveals some regional relationships with inequality.
In one study conducted in the United States, researchers looked at the country’s Gini coefficient (a measure of income inequality) and compared it to the answers of an evaluative happiness question on the GSS (General Social Survey) from 1972 — 2008 to conclude that Americans tended to be happier the years the Gini coefficient was lower (Oishi, Kesebir, and Diener 2011, 1095–1100).
Aside from socioeconomic status, we looked into a few other potential factors related to mental health and healthcare in an effort to find other potential explanatory variables for happiness. The sources for all additional variables can be found in our dataset. While none of our picks correlated as closely to the ladder score as the original six, we believe that they provide value anyway by demonstrating that often, the things we believe bring us happiness may not.
This scatter plot + trend line illustrates the relationship between the prevalence of raised blood pressure among the adult population and ladder score from 2015. The data for blood pressure is sourced from the World Health Organization’s 2020 World Health Statistics Report as a part of its Global Health Observatory (GHO) data. Unfortunately, the data is a bit dated, as our Ladder Score variable is an average from 2017-2019. Nevertheless, the biological and mental impact of blood pressure (and stress) could be interesting to look into further. The R-squared for this linear model is about ~0.364, which is fairly high.
This scatter plot + trend line illustrates the relationship between a country’s General Government Health Expenditure (GGHE-D) as a percentage of its General Government Expenditure from 2017. The data is sourced from the World Health Organization’s 2020 World Health Statistics Report as a part of its Global Health Observatory (GHO) data. The R-squared for this linear model is about ~0.347, which is also fairly high, and points us towards further research about what the relationship entails.
This scatter plot + trend line illustrates the relationship between a country’s crude suicide rate from 2016 and its current ladder score. Again, the data is relatively outdated compared to the ladder score, which is an average from 2017-2019. The data is sourced from a public data set in the World Health Organization’s Global Health Observatory (GHO) data repository. The R-squared for this linear model is about ~0.087, which is not only very low, but positive! Further research was done on this phenomenon, which follows.
The controversial conclusion of many papers researching this topic — that happier countries or states experience higher suicide rates — was tested by researchers at the University of Michigan, who instead used cities as the unit of analysis and found that the statement no longer held up (Park and Peterson 2014, 318–23).
The researchers conclude that states— never mind countries and regions — may not always be the most appropriate unit of geography. The larger the geographic measure, the less heterogeneity in experience is accounted for — which means more silences, which may have been the case of our data and the earlier researchers.
Suicide rates may a strong correlation to social comparison, as individuals compare themselves to their immediate family, friends, and neighbors when evaluating their level of happiness. Big picture analyses aren’t the best indicators for indicators like this — rather, cities and municipalities can help us gain a better understanding of individual happiness.
However, the impact of social comparison at the expense of happiness is not limited to cities or even geographic units of space. This phenomenon can also be observed in an online format, and is relevant to another key factor that deserves exploration: social media. One study highlights the negative effects that social media can have on happiness, where a correlation between social media behaviors and mental illness such as major depressive disorder (MDD) in young adults was uncovered. It was found that people who meet the criteria for MDD exhibit social media behaviors that are more mentally harmful to them, such as having a lack of friends, censoring oneself to avoid conflict, and especially, fixating on comparisons between themselves and others (Robinson et al., 2019). Clearly, in a time such as this where social media and online usage is so prevalent, this is one oversight that cannot be overlooked in order to gain a true understanding of the state of happiness in the world.
As we advocate for the importance of cultural and social influence in individual perceptions of happiness, we question the applicability of the World Happiness Report’s six factors to the diverse array of sampled nations.
We hope future research can illuminate how governments and nations as a whole may better serve their people in multiple facets of life to increase subjective wellbeing and, in turn, happiness. It seems this may ultimately be the goal of the report, given its sponsorship by the UN Sustainable Development Solutions Network, but as we scrutinize the World Happiness Report’s six variables for explaining social wellbeing, we want to ask whether that ontology dominates the discussions, measurements, and explanations of development — and thus disproportionally seeks capital growth over economic support and redistribution.