The Rust Belt to Sun Belt migration, which began in the latter half of the 20th century, continues today — despite the Sun Belt being especially vulnerable to the effects of climate change (U.S. Census Bureau, 2021). To better understand the mismatch between domestic migration and climactic livability in the United States (U.S.), we dig into the historical context around United States migration, as well as U.S. Census data analyzed through ArcGIS. *
*Maps are interactive! Please click on counties to explore details on data.
In 1950, the industrial region surrounding the Great Lakes, often referred to as the Rust Belt, was an economic giant. At the time, it accounted for 43% of all jobs in the U.S., including over half of all manufacturing jobs (Ohanian, 2014). The region would lose its strength as the decades moved on, with its share of economy-wide jobs declining by 28% from 1959 to 1980. This decline was largely attributed to the movement of manufacturing jobs overseas and towards the southern and southwestern regions of the country.
During this time period, the southern region stretching from coast-to-coast, known as the Sun Belt, experienced an economic boom. Before World War II, the Sun Belt's economy had been dominated by extractive industries (Shermer, 2013). By midcentury, many of its cities were led by business elites who prioritized attraction of industry over regulations and workers' rights. Their actions successfully attracted large corporations and employers — and employees followed. From 1940 to 1980, the Sun Belt increased its population by 112.3% (Bernard, Rice, 1983). While the Sun Belt's extreme growth in its share of the economy evened out by the turn of the Millennium, its population continues to skyrocket. Some posit that this can be attributed to the region's large growth in housing supply, which has led to relatively affordable housing compared to incomes (Glaeser and Tobio, 2007).
The below map looks at U.S Census data on net migration between states from 2016 to 2021. We see that the Rust Belt to Sun Belt migration continues today — though with a noted departure away from California and towards the Pacific Northwest. Counties receiving the highest number of domestic net migrants include Maricopa, Arizona (home of Phoenix) and Henderson, Nevada (home of Las Vegas). Cook County, Illinois, home of Chicago, has the largest number of migrants moving domestically out-of-state. The number one county they are moving to? Maricopa!
The below hotspot analysis shows "clusters" of hotspots and cold spots, respectively representing the highest gains and losses of domestic net migrants moving between states. As mentioned earlier, we see a continuance of the Rust Belt to Sun Belt migration, but with marked variations. The Midwest and North East continue to lose domestic migrants, but California and western Nevada have also joined these regions as cold spots. Three southern hotspot clusters have received a large influx of domestic net migrants moving between states: a South Eastern cluster; a Texas cluster encompassing Houston and Dallas; and a South Western cluster covering Arizona and southern Utah. There is an additonal cluster of hotspots in the Pacific Northwest, a location that is markedly distant from the Sun Belt.
The U.S. Census provides migration information on county-to-county flows over five-year periods. In exploring climate migration, the focus is on populations moving across different climates - as such, data was filtered to only include moves between states. We also want to focus on net migration flows as opposed to simply looking at incoming migration flows: while large cities such as New York and Los Angeles receive large numbers of incoming domestic migrants, they have even larger numbers of outgoing domestic migrants.
Climate scientists from multiple organizations have produced predictions on climate livability based on a variety of factors, including temperature, sea level rise, and access to water. Unfortunately, no comprehensive source on predictions for climactic livability was available on a county level. To address this, this study combines findings based on different measures by using a raster calculator in ArcGIS to create a "livability" index. Specific data used for this index includes:
Projections on a livable temperature "niche" in 2070 based on medium emmissions levels, from a paper published in the National Academy of Science (Xu et. al, 2020).
Projections on end-of-the-century mortality rates caused by climate change as a percentage of GDP under medium emissions scenarios. Published by the Rhodium Group's Climate Impact Lab, this data was only available on the state level.
Projections on water hazards based on sea level rise, flooding, and other factors, from the National Oceanic and Atmospheric Administration's Flood Mapper.
There were many data limitations in creating the 2070 climate livability index shown in the map below. The lack of county-level mortality data was a major one, as there are major differences in mortality between counties within states. There was also a lack of spatial data available on specific climate threats, including very large wildfires and "hot bulb" days in which combined heat and humidity rise to dangerous levels (Lustgarten et. al, 2020). The latter would adversely affect the livability of the west coast while the former would affect the livability of the lower midwest.
Despite limitations, enough information is available to make interesting conclusions. In general, the northern and interior portions of the country will experience the highest degree of livability in regards to climate, with the counties surrounding the Great Lakes experiencing a particular advantage. This index finds many of the most livable counties in Wisconsin, Illinois, Michigan, Vermont, Minnesota, and Colorado. It is worth noting that a more comprehensive livability index created by the Rhodium Group (and unfortunately unavailable for download) also identified many counties in Maine as highly livable, while ranking many Illinois counties as far less livable (2020).
To explore the relationship between U.S. domestic migration and climactic livability predictions, the below map shows a regression analysis between these two variables. Counties in purple indicate a strong relationship between positive net domestic migration between states and low livability predictions; counties in brown indicate a strong relationship between loss of domestic migrants between states and high livability predictions. Maricopa County shows the strongest of the former relationships and Cook County shows the strongest of the latter.
It is noteworthy that some counties, including Los Angeles and San Bernardino in California, and New York in New York state, also show a strong relationship between loss of domestic migrants and high livability predictions, despite having relatively low livability predictions. This is due to the large net loss of domestic migrants, which in turn is related to the large populaton size of these counties in general.