By: Alex Pirsos, 4/20/2020

Question

Every year, the World Happiness Report is published to rank 156 countries on citizens perception of happiness. The data considers factors such as GDP per capita, social support, life expectancy, and freedom. In 2019, The United States ranked 19th with a score of 6.882. While the report explores happiness on a global level, I want to know what contributes to happiness within The United States, and if there is an ideal geographic location within our country to maintain the highest levels of happiness.

Hypothesis

I predict that states with smaller populations and higher levels of income will rank as the happiest states. These states’ citizens will be able to live without the intense pressure of living up to high community expectations, and they will also have enough income to financially support themselves. With that said, happiness is reported to only increase up to an annual income of approximately $75,000, so this threshold will be considered when analyzing each state ( Martin, 2017 ). The aim of this analysis is to explore what determinates impact happiness on a state level.

Required Packages

R: tidyverse, ggplot2, dplyr, readr, and ggthemes

Tableau: Desktop and Public

library(tidyverse)
library(ggplot2)
library(dplyr)
library(readr)
library(ggthemes)
knitr::opts_chunk$set(echo = FALSE)

Five data sources were referenced in the preliminary section of this analysis.

read_csv("~/Downloads/happy.csv") -> USHappy
read_csv("~/Downloads/WorldHappy.csv") -> WorldHappy
read_csv("~/Downloads/WorldScores.csv") -> WorldScores
read_csv("~/Downloads/StatePopulation.csv") -> StatePopulation
read_csv("~/Downloads/State Well Being.csv") -> StateWellBeing

Set Up

Wallethub published a data set ranking the happiest states based on the following categories: emotional and physical well-being, work environment, and community environment. To start I visualized this data showing the top 10 happiest states overall and then the top ten states in each of the three categories.

USHappy %>% 
  head(10) %>% 
  ggplot(aes(reorder(State, totalScore), totalScore)) + geom_col() +
  coord_flip() +
  ylab("Happiness Score") +
  xlab("State") +
  ggtitle("Top 10 Happiest US States") +
  theme_economist_white()

USHappy %>% 
  head(10) %>% 
  ggplot(aes(reorder(State, -emotAndPhysRank), emotAndPhysRank)) + geom_point() +
  coord_flip() +
  ylab("Emotional and Physical Happiness Rank") +
  xlab("State") +
  ggtitle("Top 10 Happiest States (Emotional/Physical Health)") +
  theme_economist_white()

USHappy %>% 
  head(10) %>% 
  ggplot(aes(reorder(State, -workEnvironRank), workEnvironRank)) + geom_point() +
  coord_flip() +
  ylab("Work Environment Happiness Rank") +
  xlab("State") +
  ggtitle("Top 10 Happiest States (Work Environment)") +
  theme_economist_white()

USHappy %>% 
  head(10) %>% 
  ggplot(aes(reorder(State, -communityAndEnvironRank), communityAndEnvironRank)) + geom_point() +
  coord_flip() +
  ylab("Community Environment Happiness Rank") +
  xlab("State") +
  ggtitle("Top 10 Happiest States (Community Environment)") +
  theme_economist_white()

We see here that Hawaii, Utah, Minnesota, North Dakota, California, Idaho, Maryland, Iowa, South Dakota, and Nebraska appear in all four categories. However, after conducting my analysis, I found that there are stronger correlations for the least happy states. Below visualizes the 10 least happy states, and their rankings in the 3 categories.

USHappy %>% 
  tail(10) %>% 
  ggplot(aes(reorder(State, -totalScore), totalScore)) + geom_col() +
  coord_flip() +
  ylab("Happiness Score") +
  xlab("State") +
  ggtitle("10 Least Happy US States") +
  theme_economist_white()

USHappy %>% 
  tail(10) %>% 
  ggplot(aes(reorder(State, emotAndPhysRank), emotAndPhysRank)) + geom_point() +
  coord_flip() +
  ylab("Emotional and Physical Happiness Rank") +
  xlab("State") +
  ggtitle("10 Least Happy States (Emotional/Physical Health)") +
  theme_economist_white()

USHappy %>% 
  tail(10) %>% 
  ggplot(aes(reorder(State, workEnvironRank), workEnvironRank)) + geom_point() +
  coord_flip() +
  ylab("Work Environment Happiness Rank") +
  xlab("State") +
  ggtitle("10 Least Happy States (Work Environment)") +
  theme_economist_white()

USHappy %>% 
  tail(10) %>% 
  ggplot(aes(reorder(State, communityAndEnvironRank), communityAndEnvironRank)) + geom_point() +
  coord_flip() +
  ylab("Community Environment Happiness Rank") +
  xlab("State") +
  ggtitle("10 Least Happy States (Community Environment)") +
  theme_economist_white()

We see here that West Virginia, Arkansas, Louisiana, Alaska, Oklahoma, Alabama, Kentucky, Mississippi, Missouri, and New Mexico appear in all four categories.

Well-Being Analysis

The WalletHub happiness scores are based on 31 factors within those 3 categories. Two of those factors are career well-being and social well-being. These metrics are pulled from the Gallup Well-being Index. This index allows for a comparison of well-being scores from 2015, 2016, and 2017.

StateWellBeing %>%
  ggplot(aes(reorder(State, -Change1), Change1)) + geom_point() +
  coord_flip() +
  ylab("Change in Well-Being") +
  xlab("State") +
  ggtitle("Change in State Well-Being from 2015 to 2016") +
  theme_economist_white() +
  geom_hline(yintercept = 0, color = "cadetblue") +
  theme(axis.text.y = element_text(size = 7, angle = 0, hjust = 1))

This visualization shows the change in well-being scores from 2015 to 2016 by state. In the first year’s change, we see that only 14 states increased in well-being scores, and 2 stayed the same.

StateWellBeing %>%
  ggplot(aes(reorder(State, -Change2), Change2)) + geom_point() +
  coord_flip() +
  ylab("Change in Well-Being") +
  xlab("State") +
  ggtitle("Change in State Well-Being from 2016 to 2017") +
  theme_economist_white() +
  geom_hline(yintercept = 0, color = "cadetblue") +
  theme(axis.text.y = element_text(size = 7, angle = 0, hjust = 1))

In the second year we see a giant spike with 3 states remaining neutral and 42 states increasing in well-being. Interestingly enough, the only 5 states to decrease in the second-year comparison (Vermont, South Dakota, North Dakota, New Hampshire, and Idaho) also all decreased in the first-year comparison. Additionally, 3 of said states (Idaho, North Dakota, and South Dakota) rank in the top 10 happiest states. Perhaps sustaining high levels of happiness and well-being over time is unobtainable.

Tableau Geospatial Visualizations

In order to further visualize and compare state happiness data, I developed multiple maps in Tableau. In this section I will compare levels of state happiness to population density, household income, and crime rates. The three maps are in a workbook called State Happiness Analysis linked via the Tableau Public server.

Population Density

After analyzing happiness in correspondence to population, I determined that population alone provided misleading results because each state’s population varies drastically. Instead I analyzed population density to see if the amount of people living in proximity to you impacts one’s happiness.

In each of the following maps, the yellow circles represent states with the happiest scores (60 and over) and dark blue represents the least happy scores (below 40). As shown on the Population Density map, the happiest states typically have a low population density, but there are exceptions with California, Maryland, and Hawaii. Most of the least happy states have mid-range population densities ranging from 57.65 to 107.52. The only exception here is Alaska which has the overall smallest population density of 1.29.

There is no direct correlation here because there are states with small population densities that are unhappy and states with mid-size population densities that are happier. It is also interesting to note here that the happier states are mostly all in the Midwest and West whereas the least happy states are all in the south, except Alaska.

Household Income

Next, I looked at a map of household income to see how much money impacts state happiness. This data is built into Tableau in map layers. As stated previously psychologists have found that an individual’s happiest can increase with income up to $75,000. Once they make that much money, there are no additional increases in happiness from increased income. With that said, only Hawaii, California, and Maryland have household incomes in the highest bracket ($70,100 to $82,600), the rest of the happiest states are below this threshold.

On the opposite side we see that all of the least happy states are in the lowest income bracket ($20,100 to $51,400) with the exception of Oklahoma which is in the second lowest income bracket ($51,400 to $56,200). This shows that lower incomes have a larger negative impact on happiness that higher income have on positive impacts on happiness.

Crime Rates

Finally, we will look at crime rates to see how much crime and safety impacts the happiness of states. This visualization can be found on the Crime Rates vs. State Happiness Scores Map. First, we see that the happy states have a loose trend of having low crime rates (in the 200’s) with California and Maryland as exceptions (both in the 400s). On the flip side, we see that the least happy states mostly have high crime rates (in the 400s and 500s) with Kentucky and West Virginia as exceptions (both in the 200s).

Again, there is no direct cause and effect between crime rates and state happiness but there is a loose correlation showing that states with less crime are happier.

Assumptions

There is no one way to measure happiness. While Wallethub’s analysis and methodology is based on logic and data, happiness and its’ determinants can be interpreted with different factors. In order to follow this analysis, we need to assume that the rankings of state happiness are accurate.

Explanation

Through this exploration of state happiness and well-being, no direct causations can be drawn. The majority of happy states follow patterns including low population density, high household income, and low crime. These states are also mostly in the Midwest and West. Conversely, the majority of the least happy states have mid-range population densities, low household incomes, and high crime rates. Most of these states are also in the South.

Conclusion

While there is no direct causation between a single factor and state happiness, it appears that the conglomeration of these factors as a whole does impact state happiness. This also corresponds with Wallethub’s methodology in determining happiness as they used 31 factors to rank each state.

The US consistently ranks within the top percentile of the World Happiness Report each year. While there are major differences country to country, it does not appear that a state can dictate one’s level of happiness. Considering state income, safety, and population metrics is important when determining where to live, but while these factors can contribute to your happiness, they will not dictate it.