Following our analysis on countries’ readiness to work in a socially distant mode using digital technologies, we took a deep dive into the United States with the intent to carry out a similar analysis on the 50 states of the union. We asked how COVID-19 is impacting different parts of the US and how prepared each state is to safely begin to lift stay at home orders.

We first looked at the average weighted change in workplace, residential, and transit mobility in relation to the average change in the effective reproduction rate of the virus (R_t) from March 16 to April 26 to offer some perspective on states’ governance, in comparison to citizens’ reaction and compliance when facing COVID-19. The spread of each state on the two metrics matched with the uneven policies across the 50 states in the US.

As the Uneven State of the Union: How US States Differ in Slowing COVID-19 graph illustrates, states responding to the virus hard and early with a statewide stay at home order had more significant changes in both workplace and residential mobility and lowering R_t. In the upper right quadrant, states like California, New York, and New Jersey—where governments acted relatively fast and aggressively—had a more than 36% change in workplace and residential mobility from the baseline and fast progress in lowering the effective reproduction rate of the virus R_t.   There are, of course, states like Georgia and Texas, where state leaders were far less aggressive than peer states like California and New York in issuing and maintaining stay-at-home measures and social distancing regulations. Nevertheless, the public in these states practiced social distancing at a higher-than-average rate, as revealed by the Google Mobility data. Part of this may be due to decisions by city leaders, like Atlanta mayor Keisha Lance Bottoms, who chose speak out against premature statewide re-opening plans, and urged her city’s residents to be more cautious.

Not surprisingly, states that did not issue a statewide stay at home order, like Nebraska, North Dakota, and South Dakota are in the left bottom corner of the chart, showing less change in workplace and residential mobility and slower progress in lowering R_t. Several of these states in the bottom left corner of the chart are responsible for a considerable share of the American food supply. Americans in blue collar jobs – particularly those supporting the food supply – are exposing themselves to infection on behalf of the states that are in a better place.

Consider Grand Island, Neb., a city of just over 50,000 with a higher rate of COVID-19 cases per capita than New York City. Stay-at-home orders would have been hard in a city where a major employer is a meatpacking plant. In our analysis, Nebraska scored poorly on two crucial “COVID-cutting” measures: cutting the movement of residents (Nebraska ranked sixth from the bottom) and cutting the rate of transmission (Nebraska ranked fourth from the bottom). Unsurprisingly, Grand Island’s meatpacking plant became an infection hotspot.

Nebraska is not unique; a dozen states scored poorly on both COVID-cutting measures. Of those, four– Arkansas, South Dakota, Wisconsin and Kansas–are among the top ten states with the highest proportion of workers in food manufacturing. None of these states were successful in keeping residents at home – either because they didn’t have stringent enough orders or because residents simply didn’t have the luxury of working from home.

The states in the upper left and lower right quadrants of the plot spark several important questions: how did states manage to lower R_t despite a smaller change in mobility, and conversely, why did states with diligent social distancing have slower progress? First, both these measures are relative—all states lowered their R_t to some degree between mid-March and late April, and all states had a sizeable decrease in mobility. Second, by measuring the change in R_t, states with a low absolute R_t on March 16^th did not have as much room for progress as states with a high R_t on that same date. Finally, there are many other factors, outside of social distancing measures, which impact the spread of COVID-19.

These two primary measures serve as an evolving, imperfect view of how the disease progresses on the ground and how people respond to it during a specific period. Due to the fast-change nature and unpredictability of the virus, one might see a drastically different spread by taking a different time window.