Listen to Part 1 of the interview
Listen to Part 2 of the interview
Listen to Part 3 of the interview
Anna Rhodes is co-author of the RSF book Soaking the Middle Class: Suburban Inequality and Recovery from Disaster with Max Besbris. For two years, Rhodes and Besbris followed 59 households in Friendswood, Texas that flooded after Hurricane Harvey to better under the recovery process in well-resourced, majority-White, middle-class suburban communities impacted by natural disasters. In a new interview with the foundation, Rhodes discusses how the recovery process in middle-class communities and how it can foster inequality. The interview has been edited for length and clarity.
Anna Rhodes is Assistant Professor of Sociology at Rice University. She is the recipient of multiple RSF research grants.
Q. What motivated you to write Soaking the Middle Class? What is the “Matthew Effect” and why is it important?
My co-author, Max Besbris, and I have both studied residential decision-making in various forms in our research for a long time. So, as Harvey flooded the Houston area, we were immediately focused on asking questions about how affected households were making residential decisions in the wake of the storm. Within disaster research, significant attention had been placed on the experiences of the most vulnerable communities and residents, and rightfully so. But what that meant was that there was a gap in our understanding about how middle-class and more affluent households might be making these kinds of decisions – how they think about whether or not to return and rebuild versus sell and move away from a home that flooded during Harvey. And Harvey was a storm that was so broad, and there were actually quite a few middle-class and affluent neighborhoods that flooded during the storm. So, we wanted to jump into that gap and broaden our understanding of the way middle-class residents respond to disasters; how they would make decisions about what to do with their flooded properties - that was our initial research focus as we entered the field and conducted interviews with flooded households.
The “Matthew Effect” refers to a pattern in which those who begin with advantage accumulated even more advantage over time, while those who begin in a relative position of disadvantage become even more disadvantaged over time. The result is a growing divide between the advantaged and the disadvantaged. In our study, we see this happening during disaster recovery. Friendswood residents who were better off before the storm were those who were most likely to end up with even greater financial resources after the storm and with fully recovered homes. Those with fewer resources before the storm were less likely to have flood insurance and were more likely to have significant repairs remaining even two years later. They were often struggling to put together the finances and support necessary to fully recover from the flood. One of the biggest takeaways from our book is that even in these middle-class, well-resourced communities, which we often think of as a best case scenario for flood recovery and resilience, we still see growing inequality in the wake of disaster.
Q. Why did you choose to study the impact of Hurricane Harvey on Friendswood, Texas? Why not another climate-related disaster or another community?
Part of our focus on Harvey was born out of our own proximity to the event. We both lived through Hurricane Harvey and got to see firsthand how it affected the Houston metropolitan area. We knew we wanted to examine recovery in a middle-class community, and there were quite a few across Houston that were affected during the storm. So, in the wake of Harvey we looked at FEMA maps that were put out to illustrate areas that had been affected by floodwaters. We then looked at the median income of census tracts to try and identify communities across the metro area that were relatively close to the Houston median income. Then we went and actually drove around to look at these different communities. We found that sometimes the maps were wrong, meaning that there were very few visible signs of flooding. So, even though FEMA was taking in data and trying to assess what areas had been impacted, some areas may not have been significantly impacted or actually weren’t impacted and as data was rolling in FEMA was making corrections. In other cases, we found that the median income of the census tract was close to that of the metro area, but it seemed to be driven by a mix of more affluent single-family homeowners in one area and more affordable apartment complexes in another.
What stood out about Friendswood was that renters and owners alike were living in single family homes, so their experience of the storm and the flood itself was likely to be quite similar, as was the process of trying to recover that property. So, even if they might be in slightly different financial positions, what it would look like to navigate the flood and its aftermath was going to be relatively similar across households. That led us to focus on Friendswood as a site.
Friendswood also fits into how you would think of “middle-classness” in multiple ways. We don’t really have a single accepted definition of what it means to be middle-class in our country, but one way that we often turn to is income. But another thing that stood out about Friendswood was that it is a suburban community of single-family homes. So, it reflects that suburban ideal of homeownership as a reflection of middle-class status. It’s also a place where there was a sense of a tight knit community with robust community institutions that we think of as reflecting middle-class places. So, these things led us to settle on examining middle-class recovery and decision-making in Friendswood.
Q. Many homeowners in Friendswood considered themselves at low risk for flooding. How did they assess their flood risk before Harvey? How did this leave them unprepared?
The most common way that we saw people assess their flood risk prior to Hurricane Harvey was to use prior storms as benchmarks. The last time Friendswood saw severe flooding was in 1979, during Tropical Storm Claudette and there was some additional flooding during Tropical Storm Allison in 2001. So, when Harvey hit in 2017, these storms felt very far in the past and they served as a high watermark. People assumed that if their homes had not flooded during Claudette or Allison, then their homes were safe from future flooding. Harvey proved that that assumption was, simply, incorrect. And in the context of global warming and climate change, we will increasingly see hurricanes that have incredibly intense rainfall, like what we observed during Harvey. That means that many communities that may be considering themselves safe based on past storms are actually going to see flooding in the future.
One of the biggest consequences that we saw of this assumption of risk based on past storms was that some residents who thought their homes were at low risk for flooding chose not to purchase flood insurance. Some residents were even advised by insurance agents or others that they didn't need to purchase flood insurance if their home was in the 500-year floodplain and had never flooded before. So, the consequence of this was that in the aftermath of Harvey, those residents who didn't have insurance were left with far fewer resources to support their recovery. Whether a resident chose to purchase flood insurance was based on this assessment that if their home hadn't flooded in a prior storm, it was never going to flood. And I think that put people in a position of being unprepared for what they experienced during Harvey.
Q. The vast majority of homeowners who experienced flooding chose to stay in Friendswood after the storm and rebuild their homes. That’s not an obvious choice to make. Can you talk about the residents’ decision-making process?
We found that most residents chose to return to Friendswood and rebuild their homes in part because they had long term plans for stability in those homes. When middle-class households are purchasing homes and deciding where to live, they have the financial resources to be really intentional about that decision-making. Often people in Friendswood purchased their home with an idea that they would remain in them for the long-term. We had residents in Friendswood tell us that they planned to stay until their children finished school in the local public schools. Some planned to stay until they retired. Others planned on never leaving Friendswood and told us that they bought their homes with the idea that this was their “forever home.” So, the storm didn’t disrupt those long-term plans or change plans that were already in place. Instead, prior plans continued to orient the decision-making that people relied on in the aftermath of the storm. People were really tied to this place and wanted to continue to operate with the same kinds of plans that they had before the storm. So, we see that most people did so and chose to return and rebuilt their properties to remain in Friendswood.
Another thing that’s important to note about Friendswood is that in the aftermath of Harvey, many of the components of the social infrastructure of the community remained in place. Relatively few of the churches or schools experienced damage, and if they did it was often minor. City Hall, also, remained largely unaffected. Lots of the shops and other community amenities were still intact. So, there wasn’t a clear sense that the things that had supported the community of Friendswood would change in the aftermath.
Notably, the few households that we did see choose to move in the aftermath of the storm, typically had pre-existing plans to move – the storm just accelerated those plans. For example, we had residents who previously planned to purchase an RV and travel in their retirement. They had already been looking at RVs and deciding what kind of model they wanted to buy. Harvey just accelerated their decision to go ahead and sell their home and buy the RV. We had another respondent who was planning to leave when her oldest child was finished with high school. They decided to go ahead and made that move a year earlier than they had planned. So, for those who already had some idea that they would move and had imagined what their lives might look like in making that move, it was much easier to go ahead and activate those pre-existing plans.
Another critical piece to this is that most of our existing disaster recovery policies orient homeowners towards returning and rebuilding. People received funds through insurance, FEMA aid, or Small Business Administration (SBA) loans that are all about helping them repair their homes, none of which help them move away. We really don’t have a robust set of programmatic offerings that support people who are considering mobility or allow people the financial resources they might need in order to make that choice. There’s really only the buyout program that’s in place that could facilitate mobility away from vulnerable places and there are a ton of hoops that people have to jump through for that program. It can take months, maybe a year or longer, for households to know whether they would receive a buyout. And middle-class households don’t really have the financial resources to rent another home while they’re continuing to pay a mortgage on their flooded property and wait out the buyout process. So, we don’t really have the structures in place that would facilitate mobility away. The combination of people having intentionally bought into Friendswood, feeling tied to that community, and having clear long-term plans to live there, paired with most of the money they have access to being designed to help them rebuild and stay in place, meant that we saw most of the residents choose to stay in the aftermath of the storm.
Q. Hurricane Harvey exacerbated existing inequalities in Friendswood. Two years after the Harvey, some homeowners were in better financial positions than they were before the storm and others had incomplete repairs and found themselves burdened with large debts. How did governmental recovery policies amplify inequalities in Friendswood?
We see that policy does play a role in this disaster “Matthew Effect” that we observe in the aftermath of Harvey. One of the key factors that influenced the growing divide in financial positions of flooded households were the massive financial differences in the resources that people could access after the storm. Residents who had flood insurance were eligible to receive up to $350,000 total for both home repairs and content replacement. Households that did not have flood insurance could apply to receive FEMA aid in the aftermath of Harvey, but they were only eligible to receive $33,300 at most. That’s essentially a tenfold difference in the resources that households had access to. On average, our respondents only received around $17,000 from FEMA, which is not intended to help them fully repair their home. It’s intended to help make the house baseline safe and livable. Uninsured households were then directed to the SBA to apply for low interest disaster loans that they can use to continue repairs. But receiving a loan isn’t guaranteed. Common reasons that the SBA gives for denying loans are things like unsatisfactory credit history or repayment ability. This means that often the lowest income households are the most likely to be denied SBA loans. This reflects how policy structures can serve to amplify inequality within communities – some neighbors are going to have access to far larger recovery resources and support than others.
We then see that people broadly end up on three different trajectories in the aftermath of the storm. First, there are those who have access to flood insurance policies as well as a robust immediate response from the community that puts them on track to be fully recovered and live in a home that’s now updated to 2017 code. They are also benefiting from the quick rebound in home values in the aftermath of the storm. Within two years, homes were selling for pre-flood values, if not more. So, those who fully repaired their homes are able to benefit from that financial rebound.
The second group of households’ homes look the same as the first group. Their homes are fully repaired – you walk in and see beautiful new kitchens, etc. But they’re now paying down over $100,000 in new debt in the form an SBA loan. They might have to account for something like an additional $400 bill in their family finances each month, which, for a middle-class household, can have a pretty big impact. It can affect their monthly spending and their capacity to save and plan for the future. So, we see a second group that looks okay on the surface but has a new big financial burden as a result of their recovery.
Then there’s a third group that simply wasn’t able to put together the resources to recover. They were denied SBA loans and they received around $17,000 of FEMA assistance to begin repairing their homes. Two years later, many of these residents still had major repairs that were incomplete. They have rooms without walls. They have no flooring. They’re still working on kitchens. They may have one bathroom that’s been put back together. And many of them don’t know if or how they will ever have the financial resources to complete those repairs. They make them paycheck to paycheck and direct all of their extra funds towards trying to recover their homes. They’re in a much more precarious financial position in the aftermath of the storm. So, we see that the ways in which people have access to these different policies and resources shapes the trajectory they’re on. And all these folks can live on the same block. Driving down the street, you wouldn’t know that there are these massive differences in the financial positions that people find themselves in. It’s not until you’re inside people’s homes and talking to them about their finances that you hear about the massive differences that are emerging and the ways in which their access to different kinds of policies affects where they ended up.
Q. Aid can also come from informal sources. In the book, you talk about the local ecology of aid. What is that and how did that, in turn, worsen unequal recovery in Friendswood?
In disaster research we often talk about the importance of social networks for offering key forms of assistance to affected households. We absolutely found that social networks were key in Friendswood as the community was recovering. But, for us, the local ecology of aid is not simply the aid available from social networks, but rather the broader environment in which residents find themselves in need, which is experienced collectively. The conditions in this environment include local resources, local institutional support, social network support, and other informal forms of support. It also includes prevailing attitudes about offering, asking for, and accepting help. All of these factors shaped the ways residents approached the recovery process. And as the local ecology of aid changes, so do residents’ attitudes and approaches to asking for and accepting help. So, you’re in an environment in which there are variable resources, you’re navigating and making decisions as a homeowner who’s trying to recover from the flood, and you’re basing your decisions on what’s happening around you – that’s the local ecology of aid.
In the immediate aftermath of the storm, we saw an incredibly robust outpouring of support in Friendswood. People talked to us about not even having to ask for help. Sometimes, people were somewhat forceful in their offerings of assistance. People would actively knock-on doors asking, “What can we do? Can we help you?” And in that environment, it was quite easy for people to accept help, even in middle-class households that hadn’t typically accepted assistance before. But in this incredibly robust environment, where it seemed like everybody was offering or accepting help, it felt quite comfortable to say “Yes” to help. Need was high and offers were robust.
But we find that this fades quite quickly. Harvey hit at the end of August and by the end of November, after Thanksgiving, people talked about a massive drop off in the local ecology of aid, which creates a shift. There’s not as much of a sense that everybody is offering or accepting help. Instead, people start to worry about being stigmatized for continuing to ask for help or identifying themselves as in need. So, we found that people who were still in great need were choosing not to ask for help, even when it was available through churches or other local institutions. Sometimes they’d say no to offers of help even if they could use it. We had one respondent who reflected on this. She said her initial response to accept help was an aberration and she couldn’t believe she had accepted it. But at the time, when there were these really robust offers of assistance, it felt okay to ask people to bring her bins or set up a meal train and now she can’t believe she ever set up a meal train.
So, a shift happens. The way people are collectively experiencing the aftermath changes and offers of aid changes. Now people are thinking about the ways in which they may be judged for continuing to be in need and continuing to ask for or accept help. We find that these changes to the environment shape people’s recovery processes and might cause folks who are still in need of assistance two years later to be hesitant to ask for help. The local ecology of aid is another component of the ways in which over time, people's recovery is being impacted not just by policy, but also by the context that they're operating within.
Q. How did homeowners assess risk two years after Hurricane Harvey? What influenced their risk assessments? What was challenging about assessing flood risk after Harvey?
One of biggest takeaways we had is that even in the aftermath of Harvey people were very uncertain about how to assess their risk. Folks who, before this flood, had been certain that their homes would never flood now couldn't say that with as much certainty because their homes had just flooded. Still, people framed Harvey as an extreme or freak storm. So, they often perceived the likelihood of another or similar event as relatively low. At the same time, they know it's possible for an event like Harvey to occur again, because it has happened already. So, how to handle that uncertainty was confusing for residents.
On top of that, in the wake the storm, the city government took proactive steps and had lots of conversations about possible steps to take in order to mitigate flood risk, such as dredging Clear Creek. This added to peoples’ confusion. They’d wonder, “If my home flooded in the past, but we’ve taken these proactive steps to try and mitigate flood risk, am I at more or less risk moving forward?” Then you add climate change and people’s uncertainty about the extent to which climate change might impact hurricanes and flood risk in their community. This causes people to, again, be uncertain about to assess their risk moving forward.
The final thing that people mentioned was development. There's lots of growth and new construction happening in areas across the Houston metropolitan area. Friendswood is down the creek from much of the construction happening in Houston. So, residents were worried that that construction would also change their risk. But it was difficult to assess how much it would change their risk or what that might mean for how they should assess the likelihood of flooding again.
So, it was very challenging for individual homeowners to make sense of their risk. And they really had relatively poor information from the federal government about how they should be thinking about their risk. We have relied for many years on the idea of a 100-year or 500-year floodplain, which were essentially meaningless in the way that people were thinking about it in the context of Harvey. There have been multiple storms across the Houston metropolitan area in very short succession that are flooding homes in the 100-year floodplain, which indicates that this just isn’t an accurate assessment of risk. We need better information to help people accurately understand their risk, and to do so with a forward-looking lens that will account for the ways that climate change is going to impact their risk.
I think we’re moving in that direction. FEMA has released a new flood rating system and people are getting new information about flood risk. But we will have to wait and see about the extent to which that new information is making people feel more confident about their assessments of their risk, as well as whether it offers people clear information that will allow them to make good decisions about whether to insure their properties and how to be prepared for potential future disasters.
Q. How do growing individual disparities make a community less resilient in the face of future disasters?
We often think of resilient communities as those that have the resources and established community connections that allow the community to prepare for, respond to, withstand, and recover from a disaster or similar events. And when we observe growing disparities within a community, we see that there are an increasing number of households that are going to struggle to prepare for and recover from disasters. So, now there are many households in Friendswood that are worse off than before Harvey, and that are certainly less prepared for another storm. In fact, many of these households might struggle to put together the financial resources to weather any kind of additional shock, such as medical bill or a job loss. We've been living through a global pandemic where people are often experiencing these kinds of events.
This means that there are more households across the community that are going to struggle to prepare and recover from these kinds of events. These households make up part of the community fabric – in Friendswood and in other communities affected by disaster. And when less resourced households, in particular, continue struggling to fully recover it reflects limitations of the resilience of the community as a whole. Their struggle to fully recover means the community as a whole is less prepared for the next storm.
Q. As climate change worsens, extreme weather will only increase in scale and severity. What policy and practices would you recommend to mitigate growing inequality due to climate-related disasters?
This is a challenging question and one that many people are grappling with: How do we create policy changes that will limit the inequality that we're seeing in the wake of disasters? I think that there are a few paths forward, which we saw in the context of our research. One is that we must do better to help households accurately understand their risk. This will allow them to make effective choices about their insurance needs and other needs, such as where to buy or build homes. This will help us reduce the impact of disasters and allow people to have the resources required to recover should they experience disaster in the future. Then, we would ideally see fewer disparities between households in their preparation for these kinds of disasters, such as getting flood insurance, since they would have a more accurate understanding of their risk.
Another thing we saw in the aftermath of Harvey was that there were a wide range of experiences that people had working with their mortgage lenders to try to get a break on mortgage payment schedules. The COVID eviction moratorium has highlighted that allowing households in financial distress time to recover can be extremely beneficial. We don't have a similar requirement or set of standardized practices in place in the wake of disaster. We should think about developing standard policies around how mortgage companies might provide forbearance from payments, and potentially a foreclosure moratorium for mortgage holders. This could allow residents some time to make decisions and put together the resources that they need to be able to recover. It would also hopefully reduce some of the differences in resources that households would be able to put towards recovering their properties. That would also hopefully reduce the likelihood that financial stress drives choices immediately after disaster. It might allow for some breathing room to decide whether moving away from that vulnerable place is actually a better path than staying and rebuilding. It could allow folks time to determine whether to sell their property or what that might look like.
I think one of the biggest things that we need to do is to broadly facilitate an expansion of policies that allow for mobility away from risky places because we see inequality emerging and growing as people try to return and rebuild in place. There may be pathways forward that involves supporting strategies of managed retreat from these locations that will reduce people's future risk and allow them to navigate recovery without continuing to face the potential that their homes will flood in the future. We really don't have a robust set of policies that facilitates that kind of mobility away. So, I’d recommend expanding the buyout program, developing new programs, increasing funding, and changing eligibility requirements. Sometimes households were open to potentially accepting a buyout but found out that they actually weren't eligible, even though their home had flooded during Harvey. So, we need to think about how we're drawing those eligibility requirements, what those requirements mean, and investing in the idea that we should support mobility away from the most vulnerable places as a key part of our policy strategy. Responding to climate change is going to be something that we need to continue to focus on moving forward.
Finally, it was notable to me that within this middle-class place, that while there were many households that we're still struggling two years after Harvey, it seemed like people across the community were largely unaware that that was true. So, we should think about strategies for identifying households that are continuing to struggle to recover over time. That way, we can effectively target resources to support recovery towards households that continue to need it. That will help address the issues around the resilience of the community as a whole. If everyone is able to recover fully, then the community can really think about preparation for the next storm; they won’t have to continue navigating recovery processes from the last storm. Thinking about resilience and thinking about targeted, ongoing, long-lasting policy that will help everyone recover will be key.