To determine America’s 50 worst cities to live in, 24/7 Wall St. considered the roughly 600 cities that the U.S. Census Bureau reported as having populations of more than 50,000 in 2016. Data was collected in nine major categories: crime, demography, economy, education, environment, health, housing, infrastructure, and leisure.
Within each category, specific measures contributed to a city’s overall category score. For example, the economy category included median household income adjusted for cost of living, the ratio between a city’s and its state’s median household income, poverty and unemployment rates, as well as a city’s three-year employment growth. Each measure was adjusted to range from 0 to 1 using min-max normalization, with lower scores indicating better outcomes.
Normalizing each measure, as opposed to aggregating category scores in other ways, allowed us to assign weight to individual measures for added importance rather than entire categories. It also enabled us to expose the principal components of our index — those measures with wider variation that disproportionately determine the rank of a city’s composite score. The housing category, for example, had the widest range, giving it the greatest pull in our index. Crime and economy also had large variances.
We did not include any measures in the demography category in our composite index. However, this category provided exclusion rules. Cities that are better to live in often attract job seekers and their families. Conversely, labor market slack, unaffordable housing, high crime rates, or other negative factors may induce people to move to a different city with better prospects. Population figures are from the Census Bureau’s 2016 American Community Survey (ACS).
The crime category consists of both violent and property crime rates from the Federal Bureau of Investigation’s 2016 Uniform Crime Report. High crime rates have the potential to make a city less desirable. As a result, cities with crime rates lower than the national rates were rewarded, while cities with higher crime rates were penalized.
The economy category includes a city’s 2016 unemployment rate and employment growth from 2014-2016, both from the Bureau of Labor Statistics. Cities were penalized for having poverty rates above the national rate of 14.0%. Our goal was to identify cities that were livable for everyone, not just the rich. Still, if incomes are too low, a city may not be desirable. To that end, we adjusted median household income for cost of living in the city. Cities were penalized if cost-adjusted incomes were less than $44,000 or more than $112,000 a year, roughly 80% to 200% of a typical household’s income nationwide. Poverty rates and median income came from the 2016 ACS. Cost of living data came from property and real estate data provider ATTOM Data Solutions.
As a proxy for school system strength, we considered high school standardized test scores relative to state scores from ATTOM Data Solutions. Test score data are for 2015 or the most recently available year. Additionally, the education category included the percentage of adults with at least a bachelor’s degree from the ACS, as well as the number of colleges and universities in a city per 100,000 residents from the Department of Education. Both measures are as of 2016.
In the environment category, we included an air quality index to assess the levels of a variety of pollutants on a given day. Additionally, we considered an index measuring natural hazard risk, as well as average monthly rainfall. All data in this category came from ATTOM Data Solutions.
From the Centers for Medicare and Medicaid Services (CMS), we calculated 30-day risk-adjusted mortality rates of heart attacks, COPD, heart failure, pneumonia, and stroke. Also from CMS, we examined the rate at which individuals were readmitted to a hospital within 30 days of being discharged. This category also includes preventable hospitalizations — the share of hospitalizations that could have been treated with outpatient or ambulatory care for every 1,000 Medicare recipients from County Health Rankings.
In our housing index, we considered the ratio of a city’s median home value to the statewide median home value. Cities were penalized if a city’s home values were worth less than 90% of statewide home values. Conversely, if home values were typically 25% higher than across the state, high barriers to entry exist that can make a city unaffordable. As an additional measure of affordability, we included the ratio of median home value to median annual household income. This ratio — called a price-to-income ratio — helps identify cities that are less liveable for a broad audience. We also considered median property taxes as a percentage of median home value. All data in this category came from the 2016 ACS.
From the ACS, we considered the percentage of commuters travelling to work by foot or public transportation. Additionally, we reviewed the average time it takes to travel to work each day. Lastly, we included the number of airports in the metro area in which the city is located. There are, for example, no airports in New York County, the primary county of New York City. However, at least three major airports exist outside county limits — and within the metro area — that service people who live in the city. Airport data came from the Federal Aviation Administration and only consider operational public-use and commercial airports as of 2015.
The leisure category can be broken into two parts: activities that take place in the city and outside it. Within a city, residents may take advantage of restaurants and bars, libraries and archives, theater companies, fitness and recreational sports centers, museums, movie theaters, hotels, or amateur and professional sports teams. To engage in other pastimes — skiing, for example — residents likely have to leave city limits. We included in this index the number of zoos, nature parks, ski resorts, and golf courses in the counties surrounding the city. All data in this category were aggregated to the city level from 2015 Zip Code Business Patterns, a program maintained by the Census and adjusted for the city’s 2016 population.