How Much Do People Attribute their Happiness to the Economy?
The Super Statisticians
Monday, December 17, 2018
1 Introduction
Currently, the standard measure of a country’s success is in terms of GDP. This monetary perspective, however, is limiting, and not a complete view. Thus, another aspect of that could provide insight on prosperity is happiness, and by bridging economy and happiness, the two measurements create a more comprehensive view of the definition of success. Happiness is now a valid measurement as more and more government institutions and organizations are using happiness data to inform policy-making. This is important as it reveals that at the core of each policy, there is an innate value of happiness and not an institution’s fiscal interests. This moral consideration of policies hopefully will make for more compassionate and effective governments and a generally “happier” world.
Our dataset uses answers collected from the 2017 Gallup World Poll in which participants were asked to use what is known as the “Cantril Ladder” to assess and rank various aspects of their lives. The assessment is as follows: a score of 10 correlates to the best possible life while a score of 0 correlates to the worst possible life. Participants were then asked several other questions pertaining to economic production, social support, life expectancy, freedom, absence of corruption, and generosity and asked to rate how much of an impact these factors had in their quality of life.
Research Question
How much do people attribute their happiness to economic prosperity?
Data Interpretation
Our research looks at the economic factor and the happiness score, a relationship that tells us to what extent people contribute their happiness to economic prosperity. Thus, this is not a simple correlation between amount of money and happiness, but a correlation between perceived happiness and the associated perceived attribution to the economy. This relationship is nevertheless still important for the aforementioned reasons that pertained to happiness and GDP.
2 Exploratory data analysis
| continent | correlation |
|---|---|
| Africa | 0.596 |
| Americas | 0.765 |
| Asia | 0.758 |
| Europe | 0.809 |
| Oceania | -1.000 |
Observations About Preliminary Visualizations:
Our initial visualizations allow us to quickly see correlations and trends. It is clear that the two histograms appear to overlap which already suggests a strong correlation between the explanatory and predictor variables.
All continents but Oceania have positive correlations and slopes. As a result, all continents except one can be described as having people where the more they attribute economic growth to their happiness, the happier they are. Oceania is an exception where we see the reverse: the more people attribute economic growth to their happiness, the less happy they are. Since there are only two data points for Oceania, this is a conclusion that is not as concrete as others. Nonetheless, there is still a negative slope and correlation.
3 Multiple regression
We modeled our multiple regression using the parallel slopes model, which consisted of one numerical explanatory variable, economy, as a factor of GDP Per Capita, and one categorical explanatory variable, continent. The numerical outcome variable is the happiness score.
| term | estimate | std_error | statistic | p_value | lower_ci | upper_ci |
|---|---|---|---|---|---|---|
| intercept | 3.204 | 0.126 | 25.436 | 0.000 | 2.955 | 3.453 |
| Economy..GDP.per.Capita. | 1.849 | 0.157 | 11.752 | 0.000 | 1.538 | 2.160 |
| continentAmericas | 0.913 | 0.170 | 5.374 | 0.000 | 0.578 | 1.249 |
| continentAsia | 0.173 | 0.150 | 1.154 | 0.250 | -0.123 | 0.470 |
| continentEurope | 0.441 | 0.173 | 2.549 | 0.012 | 0.099 | 0.782 |
| continentOceania | 1.423 | 0.454 | 3.136 | 0.002 | 0.526 | 2.319 |
3.1 Statistical interpretation
- Intercept: 3.204 is the intercept of the baseline group Africa and represents the mean happiness score
- Economy..GDP.per.Capita.: 1.849 is the common slope for all lines. For every increase in 1 unit of the economy variable there is an associated increase of 1.836 in happiness score.
- continentAmericas: 0.913 is the offset of the intercept for the continent Americas relative to Africa. The intercept for the Americas group = 2.762
- continentAsia: 0.173 is the offset of the intercept for the continent Asia relative to Africa. The intercept for the Asia group = 2.022
- continentEurope: 0.441 is the offset of the intercept for the continent Europe relative to Africa. The intercept for the Europe group = 2.29
- continentOceania: 1.423 is the offset of the intercept for the continent Oceania relative to Africa. The intercept for the Oceania group = 3.272
The modeling equation for the above scenario is:
\(\widehat{y}\) = \(b\)0 + \(b\)1\(*\)\(x\)1 + \(b\)2\(*\)\(x\)2 + \(b\)3\(*\)\(x\)3 +\(b\)4\(*\)\(x\)4 + \(b\)5\(*\)\(x\)5
The modeling equation for the Americas:
\(\widehat{Happiness Score}\) = 3.204 + 1.849\(x\)1 + 0.913*1Americas(\(x\))
The parallel slopes model is similar to the exploratory data visualization in that they both show positive slopes for the continents with the exception of Oceania which has two data points.
As mentioned in the beginning of this report, the happiness score was obtained and analyzed across more than an economic factor. It is important to recognize that this happiness score is not purely from the economic growth perspective, but many other considerations as well. So in our analysis it’s important to recognize the relationships between the economic variable and happiness score but in no means stating that the economic variable is the only contributing factor and in our case, increased happiness.
3.2 Non-statistical interpretation
- Across all continents it is observed that people attribute a good quality life, and their happiness, to economic growth. When economic growth had a smaller impact on their lives, happiness had lower scores. Again, it is important to remember that there are other factors affecting the score.
4 Inference for multiple regression
When looking at the regression table, we can conclude whether or not the data is suggestive that there is or there isn’t a relationship. If the “no relationship” value of “0” is in the confidence interval, then the conclusion can be made that there is not a relationship between how people attribute economic prosperity to their happiness.
For our p-value analysis we are using \(\alpha = 0.05\) because this is a value that is most often used in literature as well as not being extremely liberal or conservative. If the p-value is less than \(\alpha\) we reject the null hypothesis \(H\)0 which states there is no relationship between how people attribute economic prosperity to their happiness. We then accept the alternate hypothesis (\(H\)A) that there is a relationship. If the p-value is greater than \(\alpha\) we do not reject the null hypothesis. This means that \(H\)0 may still be true but there is not enough evidence within the data to reject \(H\)0.
When looking at Africa’s confidence intervals of [1.583, 2.160] the data is suggestive that there is a relationship between how people attribute economic prosperity to their happiness. Since the p-value is 0.000 and less than \(\alpha = 0.05\), we reject the null hypothesis and conclude there is a relationship.
When looking at Americas’ confidence intervals of [0.578, 1.249] the data is suggestive that there is a relationship between how people attribute economic prosperity to their happiness. Since the p-value is 0.000 and less than \(\alpha = 0.05\), we reject the null hypothesis and conclude there is a relationship.
When looking at Asia’s confidence intervals of [-0.123, 0.470] the data is suggestive that there is not a relationship between how people attribute economic prosperity to their happiness. Since the p-value is 0.250 and greater than \(\alpha = 0.05\), we do not reject the null hypothesis and conclude there is a possibility of no relationship.
When looking at Europe’s confidence intervals of [0.099, 0.782] the data is suggestive that there is a relationship between how people attribute economic prosperity to their happiness. Since the p-value is 0.012 and less than \(\alpha = 0.05\), we reject the null hypothesis and conclude there is a relationship.
When looking at Oceania’s confidence intervals of [0.526, 2.319] the data is suggestive that there is a relationship between how people attribute economic prosperity to their happiness. Since the p-value is 0.002 and less than \(\alpha = 0.05\), we reject the null hypothesis and conclude there is a relationship.
- The histograms have a normal distribution and do not have an extreme skew. In the scatterplot, the residuals have constant variance/spread with respect to the economy factor. Thus, all the assumptions for inference are met so all of the p-values and confidence intervals in the regression are valid.
5 Conclusion
In our analysis, we tried to determine if people attribute their happiness to economic prosperity. Our analysis indicates there is a positive correlation between the economic factor and happiness, meaning when people attribute more of their happiness to economic prosperity they are in return happier.
For every continent except Asia we rejected the null hypothesis thus there was a suggested relationship. Since we did not reject the null hypothesis for Asia we concluded that there was a possibility of no relationship, however the null hypothesis may still be true but there is not sufficient evidence to reject it. The residual analysis confirmed the assumptions made in these inferences since all the conditions were met.
There are a couple limitations to our data. The first is the lack of information regarding the population’s economic status while answering this happiness survey. It would be interesting to know possible differences of answers between those of lower, middle, and high economic class. Another limitation is the possibility of an inaccurate self-score since there is an emotional and subjective element that plays into assessing one’s happiness. For example, if someone is upset or particularly emtotional while assessing themselves, it is possible that the overall happiness score would be different. But due to the nature of the test, there is no way to get objective results.
This report takes a look at the relationship between happiness and the economy but through a more psychological approach. Thus we looked at the economy factor, a variable used to describe the extent GDP contributed to their happiness. The outcome variable, happiness score, takes into account various factors and not just the economic one to come up with an overall value for their measured happiness. In future work, it would be interesting to examine social support, life expectancy, freedom, absence of corruption, or generosity since these are factors that people also attribute to happiness.
6 Citations and References
Sustainable Development Solutions Network. “World Happiness Report.” RSNA Pneumonia Detection Challenge | Kaggle, 14 June 2017, www.kaggle.com/unsdsn/world-happiness.
Baggini, Julian. “Forget GDP – There’s More to Britain’s Wealth than Its Bank Balance | Julian Baggini.” The Guardian, Guardian News and Media, 14 Aug. 2017, www.theguardian.com/commentisfree/2017/aug/14/gdp-measure-progress-housing-healthcare-technology#img-1.