Andrey Kamenov, Ph.D. Probability and Statistics
Have you ever seen “quality of life” comparisons between different countries? There are many ways to measure such things (some involving much more guesswork than the others). Among the most prominent is the Economist’s where-to-be-born index, which we have thoroughly discussed in previous posts.
There are many fewer studies on how quality of life differs between the states in our country. Moreover, this is not an empty question: California alone could be the fifth-largest economy in the world. Wouldn’t it be logical to assume that life in North Dakota may be somewhat different?
Of course, there are some issues we should deal with before we can make such comparisons. Some variables, like those describing the state of political freedoms, are meaningless in our state-to-state comparison. Others, like gender equality, don’t exhibit any statistically meaningful variation.
Another issue lies in weighing the factors. The most common way to do this (employed most notably by the Economist’s index mentioned above) uses regression techniques to fit life satisfaction survey data. This approach measures the importance of each factor by its impact on life satisfaction scores.
The problem is, the survey data is not exactly reliable. Take a look, for instance, at the following chart comparing the Gallup-Healthways Well-Being Index and OECD Better Life Index scores for all 50 states:
Sure, there is indeed some positive correlation between the scores, but it is not exactly strong. There are states that place low on one list and high on another: Montana and Alaska present two obvious examples.
During our research, we found that Gallup survey scores correlate more significantly with the objective factors we were using. Therefore, we’ll stick to this survey for the rest of this post.
So, what factors did we use? We combined economic (household disposable income per capita and unemployment rate), social (homicide and divorce rates and life expectancy) and climate (air pollution and general weather comfort, based on an earlier post) factors.
We concluded that factors related to climate and life expectancy mattered the most. Counterintuitively, the impact of disposable income on life satisfaction was lower than expected.
To conclude, here is a map showing both actual and fitted values, as well as how the states rank according to each one of our seven factors.
Source(s):
About Andrey Kamenov
Andrey Kamenov, Ph.D. Probability and Statistics
Andrey Kamenov is a data scientist working for Advameg Inc. His background includes teaching statistics, stochastic processes and financial mathematics in Moscow State University and working for a hedge fund. His academic interests range from statistical data analysis to optimal stopping theory. Andrey also enjoys his hobbies of photography, reading and powerlifting.
Other posts by Andrey Kamenov: