The U.S. Gulf Coast has been hit hard this hurricane season. First came the devastating Hurricane Ida, which washed away homes and knocked out power for more than a week along the Gulf Coast. Then came Hurricane Nicholas with heavy rains, further slowing down the already slow efforts at recovery.

Who’s likely to be helped first? When disasters strike, governments and utility companies claim they help everyone equally, ignoring race and socioeconomic status. But on the ground, many argue that poorer and less well-connected neighborhoods are not only more likely to be harmed by those disasters, but will also wait longer for government assistance and resources.

A substantial amount of research finds that those suspicions are accurate. Socioeconomic conditions like race, class and gender influence which communities are hardest hit and how quickly governments respond.

That matters. Promptly restoring power and quickly delivering disaster relief can save lives. Outages hurt health in numerous ways, resulting in medical supply shortages, weakened health systems and even loss of life for those who rely on medical devices like ventilators and dialyses.

But even though some communities are more vulnerable to disaster to begin with, utilities tend to ignore those demographics when restoring electricity, which we call the “colorblind restoration approach.” That can be a problem. As a result, these companies may unintentionally restore service more quickly in areas that are more socioeconomically advantaged than others. Being “colorblind,” in other words, can discriminate without being designed to do so.

Lessons from the 2017 hurricane season

As the climate changes, tropical storms and hurricanes are growing more frequent and intense. With marginalized communities already on the front lines for hurricane damage, that means that unless the international community and the United States adopt policies to curb climate change and mitigate its effects, those communities will be increasingly hit by disasters.

That’s what our research found when we studied the 2017 hurricane season, the costliest to date and one of the deadliest in U.S. history. In just one month, Texas, Florida and Puerto Rico were hit by some of the strongest hurricanes in the history of Atlantic tropical storms. Hurricane Maria devastated Puerto Rico and its infrastructure, destroying key electricity transmission and distribution lines and eventually becoming the largest blackout in the history of any U.S. state or territory, as measured in total customer-hours of lost electricity. Hurricane Maria left the Puerto Rican archipelago without electricity, offering us a unique chance to investigate a key question: When everyone loses power, who gets power back first?

Marginalized groups wait longer for power restoration

The Hurricane Maria power restoration efforts featured broad disparities. While some communities received power restoration crews within 12 days, others waited more than 300 days for crews to arrive to their communities. On the ground, marginalized communities wondered whether they would have to wait longer while privileged communities got their power back quickly. Three months after Hurricane Maria, mayors from the opposition party (PPD) lodged formal complaints with federal agencies and led a rally, objecting to what they claimed were political biases in restoring energy — with power going first to municipalities led by the ruling Partido Nuevo Progresista (PNP) party. Some researchers have found that elected officials direct disaster resources to reward voters for their electoral support. But was that happening in Puerto Rico?

To examine whether the mayors’ allegations were accurate, we requested data from the Puerto Rico Electric Power Authority so that we could see when and where restoration crews were deployed after Hurricane Maria. We also interviewed key electric utility company officials involved in the recovery effort; we are keeping their names and positions confidential to prevent potential retaliation. All the officials we interviewed claimed that in assigning restoration crews, they followed priorities set down before the hurricane, putting critical infrastructure first, including hospitals and emergency operations centers.

However, our statistical analysis found that even accounting for these priorities as well as other factors impacting the recovery, such as hurricane damage and economic priorities, the mayors were correct: Government did restore electricity faster to people living in areas that had supported the PNP. What’s more, communities with greater numbers of socially vulnerable people waited longer than more privileged communities for power restoration crews to reach them — despite the fact that socially vulnerable communities tend to support the PNP in greater numbers. Opposition party strongholds waited still longer, whether they were especially hard-hit by the hurricane or were more remote and therefore harder to reach.

We term existing approaches to energy restoration “colorblind restoration” because they purposefully ignore community demographics when deciding where to send power restoration crews. Our research demonstrates that ignoring community demographics this way can reproduce existing inequalities without necessarily having the intention to do so.

Power restoration can increase inequality

This inequality in recovery, which prioritized some communities while neglecting others, was worsened by previous inequality in disaster preparation and compounded with the federal government’s ineffective and unequal disaster response in U.S. states during the 2017 hurricane season. Socially vulnerable communities tend to lack the resources needed to prepare for disasters. Disaster preparedness spending in the United States is often used to serve electoral aims, with resources disproportionately allocated to counties likely to vote for the party of the incumbent president. In Puerto Rico, the austerity measures imposed by an unelected and highly unpopular Fiscal Oversight Board led to a drastic reduction of utility workforce.

Inequalities in vulnerability to disasters occur not just in Puerto Rico but throughout the United States. In an Axios-Ipsos poll conducted earlier this month, Black Americans were more likely to report that they’d endured power outages in the previous year. When these outages combine with extreme weather — for instance, heat waves after Hurricane Ida or freezes during the Texas winter storm in February — consequences can be catastrophic.

Failing to make vulnerable communities a priority in preparing for and recovering from disasters means that more people are likely to die — and poorer communities of color are the most likely to suffer.

Fernando Tormos-Aponte (@fernandotormos) is an assistant professor at the University of Maryland Baltimore County, a Union of Concerned Scientists Kendall fellow, and a visiting scholar at Johns Hopkins University.

Gustavo García-López is an assistant researcher at the Centre for Social Studies at the University of Coimbra and a 2019-2021 Prince Claus Chair at the International Institute of Social Studies at Erasmus University Rotterdam in the Netherlands.

Mary Angelica Painter (@MaryAPainter) has a PhD in political science from the University of Missouri at St. Louis and is a director of the Confluence chapter of the Scholars Strategy Network.