|Home | About | Journals | Submit | Contact Us | Français|
This article introduces us to the complexities of conducting social scientific research in a major urban disaster zone and reports on results from the most systematic survey at the time aimed at tracking the whereabouts of pre-Katrina residents during the first year of recovery, the Displaced New Orleans Residents Pilot Survey. This survey drew an area-probability sample of pre-Katrina residents and set out to interview them approximately twelve months after the storm, when educated guesses placed the New Orleans’s city population at roughly half its pre-Katrina total. Results confirm that early returnees tended to be disproportionately white, elderly, better educated, and far less likely to have homes rendered uninhabitable by the disaster. These patterns begin to show how demographic processes triggered by the disaster exacerbated existing inequalities in the region, allowing more advantaged residents to return while leaving less advantaged residents dispersed across numerous destinations.
Hurricane Katrina displaced virtually the entire population of New Orleans, scattering people throughout the region, state, and country. In the months following the disaster, there was no source of representative information on people’s whereabouts, well-being, or their plans to settle in a new location or return to the city. We designed and fielded the Displaced New Orleans Residents Pilot Study (DNORPS) to provide preliminary information on these topics and to assess the feasibility of collecting representative data on this population. The pilot study was based on a probability sample of pre-Katrina dwellings in New Orleans and was fielded in the fall of 2006, approximately one year after Hurricane Katrina. It was designed to form the basis for a larger, more extensive effort to track and interview displaced New Orleans residents that is currently underway. The information from a full-scale longitudinal survey will help us understand the effects of Hurricane Katrina over the medium and long term and to investigate how individual, family, and contextual characteristics shaped the population response to this environmental disaster (Briggs 2006; National Academy of Sciences 2007).
The data from DNORPS are of particular value for studying the recovery of New Orleans. The DNORPS provides one of the few sources of information on displaced residents who have not returned to New Orleans, which allows comparisons between the pre-Katrina residents of the city who have and have not returned. This comparison provides valuable insights into key factors that shape decisions about whether to return to the city. With DNORPS data we can also gain insights into the future population of New Orleans from estimates among returned residents of their likelihood of staying in the city and among displaced residents of their likelihood of returning.
In this article, I show that in the fall of 2006 approximately one-half of residents had returned to New Orleans, and by the fall of 2007 I estimate that two-thirds will have returned. The most important factor associated with residents’ return to the city is having a habitable pre-Katrina home. Living in an unflooded or moderately flooded part of the city is also positively associated with return. Those who have returned are older, disproportionately white, more highly-educated, and more likely to be employed than those who are still displaced. Displaced residents who expect to be living in New Orleans within the year are more often those who rented their pre-Katrina dwellings. The survey methods applied to obtain these results demonstrate that, although difficult, it is possible to successfully survey a displaced population. Furthermore, they reveal some of the selective social processes shaping return migration.
Studying population change in New Orleans after Hurricane Katrina has been a major challenge because valid and generalizable results require data that are representative of everyone and outcome measures that are relevant for describing the post-hurricane experience. The difficulty is that some population groups are difficult to find. For example, it is harder to track and interview those who have resettled away from New Orleans. At the same time, national studies that include displaced New Orleans residents may not track relevant outcomes and may only provide a brief window for examining their circumstances.
Data sources for studying the demographic effects of Hurricane Katrina fall into two categories. The first is early and generally small-scale studies of evacuees. The second is larger-scale surveys, including ongoing national surveys. Researchers conducted early studies rapidly and under very demanding circumstances. Although these studies provide useful information on the immediate aftermath of Hurricane Katrina, they focus on special populations, such as evacuees in shelters or families that had registered with the Red Cross. They were also exclusively cross-sectional and often lacked representative samples owing, for example, to unknown but possibly high non-response rates and various shortcomings of their sampling frames (see, for example, Abramson & Garfield 2006; Brodie, Weltzien, Altman, Blendon & Benson 2006; Elliott & Pais 2006; Stone, Grant & Weaver 2006). However, early information on the current location of evacuees was also available from change-of-address forms filed with the U.S. Postal Service and from registrations for aid with the Federal Emergency Management Agency (FEMA). Analyses of these data show, for instance, that nearly 15 percent of evacuees from New Orleans relocated to distant cities in the East Coast, Midwest, and West Coast (Tizon & Smith 2005).
Larger-scale and longer-term studies are likely to provide key insights into the location and well-being of displaced New Orleans residents. These studies generally have well-designed sampling plans and, for studies focusing on the effects of Hurricane Katrina, they include relevant outcome measures. Among the studies that we review are the Hurricane Katrina Community Advisory Group, the Louisiana Health and Population Survey, the Kaiser Post-Katrina Baseline Survey, the Current Population Survey, and the American Community Survey.
The Hurricane Katrina Advisory Group (2006) is a panel study of 1,043 adults from Alabama, Mississippi, and Louisiana who were selected from multiple sampling frames. This study focuses on the psychological well-being of adults. Although this was the first large-scale study launched after Katrina, it suffers from a number of shortcomings for addressing topics of interest to social scientists, such as migration and broader measures of well-being. Among these shortcomings are a response rate of approximately 27 percent, no information on characteristics of non-respondents, a total of only 168 respondents in New Orleans, and no contextual information on respondents’ pre-Katrina neighborhood of residence.
The Louisiana Population and Health Survey (2006) collected cross-sectional data for parishes in Louisiana that were directly affected by Hurricane Katrina, including Orleans Parish. Researchers collected basic demographic information for all respondents in currently occupied households as well as some broader measures of health status and social and economic well-being. Similarly, the 2006 Kaiser Post-Katrina Baseline Survey (Kaiser Family Foundation 2007) is a cross-sectional survey that interviewed 1,504 randomly selected adults living in the New Orleans metropolitan area, including 901 respondents in Orleans Parish. It collected information on several dimensions of well-being, including economic status, physical and mental health, and quality of life. These two studies provide complementary information to DNORPS. All three studies provide information on individuals who lived in New Orleans prior to Katrina and have subsequently returned to the city. Only DNORPS collects information on New Orleans residents who have not returned, while the Louisiana Population and Health Survey and the Kaiser Post-Katrina Baseline Survey collect data on new migrants to the city.
Finally, the American Community Survey (ACS) and the Current Population Survey (CPS) provide valuable data on the entire region affected by Katrina (U.S. Census Bureau 2006; Bureau of Labor Statistics 2006). These national surveys are fielded by the U.S. Census Bureau and achieve high response rates. Sample sizes in the CPS are too small to provide detailed estimates for New Orleans, but sample sizes in the ACS are sufficiently large to identify and analyze displaced New Orleanians as well as the current population of the city. Both studies will provide only a brief window into the whereabouts and well-being of displaced New Orleans residents: they each ask just a single migration question about place of residence one year previously, and the CPS special questions on Katrina migrants ended in October 2006 (Cahoon et al. 2006).
The Displaced New Orleans Residents Pilot Study was planned to provide key information and insights for launching a major new panel study of displaced New Orleans residents. In this section, we describe DNORPS, providing an overview of the study design, questionnaire content, fieldwork operations, and fieldwork results
DNORPS is based on a stratified, area-based probability sample of pre-Katrina dwellings in the City of New Orleans. In area-based sampling, researchers select dwelling units at random from the entire geographic area of the study in a single stage leading to more precise estimates than alternative sampling designs. In other words, dwellings were the primary sampling unit and there was no geographic clustering of cases. We sampled 344 cases for the study, with the sample size determined by the budget. We selected these cases at random from two separate lists of housing units, a “named” list that included the name of the head of household and an “unnamed” list. The sample was also stratified by flood depth. Using DNORPS sample weights, the data provide representative information on pre-Katrina residents of New Orleans.
We divided New Orleans into three strata (see Figure 1) on the basis of flood depth on August 31, 2005, for sampling purposes. Flood depth was directly tied to housing damage (McCarthy, Peterson, Sastry, and Pollard 2006), and thus represents a crucial factor shaping residents’ migration decisions (such as whether to return to the city) and financial and other dimensions of well-being. The first stratum was areas with no flooding, which included about 29 percent of all dwellings in New Orleans. The second stratum included areas with less than four feet of flooding, or about 20 percent of all dwellings. Finally, areas with four or more feet of flooding made up the third stratum, which included 51 percent of dwellings.
We drew the sample from a listing of all dwellings in the city of New Orleans created by Marketing Systems Group, Inc. (MSG), which develops and maintains an enhanced listing of all households in the United States on the basis of records from the U.S. Postal Service and other sources. We used an archived version from August 2005 to draw the sample. We oversampled those dwellings for which MSG provided a name of the household head, under the assumption that these cases would be easier to locate. For similar reasons, we oversampled dwellings in the no-flood and low-flood strata and undersampled dwellings in the high-flooding strata. The oversampling rate was modest, however, and sample weights can be used to generate estimates for the entire population.
The DNORPS questionnaire obtained a complete roster of all pre-Katrina household residents, their relationships with one another and their background characteristics, such as age, sex, race and ethnicity, place of birth, and education. The questionnaire collected information about the evacuation and resettlement experience, including when each person left New Orleans, the current place of residence, and, if they currently resided in New Orleans, when they had moved back to the city, as well as the likelihood that each person would be living in New Orleans in one year’s time. The questionnaire asked the respondent to report about each household resident’s pre-Katrina and current employment, marital, and health status. Information collected on housing characteristics includes the type of dwelling, whether the unit was owned or rented, and the extent of damage from Hurricane Katrina and the subsequent flooding. It took respondents approximately fifteen minutes, on average, to complete the ten-page questionnaire. Respondents received a check for $20.00 as an incentive for their participation.
Fieldwork for the DNORPS began in mid-September 2006 and ended in November 2006. The goal was to complete interviews with an adult respondent in each sampled household through either a mail survey, a telephone interview for those who did not complete the mail survey, or an in-person interview for respondents who were unable or unwilling to complete the survey by mail or telephone.
We first mailed a cover letter and questionnaire to all 344 cases, with an address correction requested. As questionnaires were mailed out, we began batch tracing of respondents using electronic databases such as Telematch, TransUnion, Death Audit Search, National Change of Address, and Accurint to obtain names (for the unnamed sample) and up-to-date telephone numbers and addresses. We obtained names for 270 cases (78 percent of total) and telephone numbers for 230 of these cases, leaving 114 cases to be traced interactively by telephone or in person. We sent cases with a telephone number to the call center for an attempt at a telephone interview. Cases without a telephone number, or which could not be reached by telephone, were sent to one of four field interviewers located in New Orleans for on-the-ground tracing. Fieldwork ended in mid-November 2006 when the budget was exhausted. At that time, 113 cases had pending tracing activities because we had been unable to confirm a name, telephone number, or new address and had not determined the dwelling was vacant at the time Hurricane Katrina struck.
A descriptive analysis of the fieldwork results is presented here. Elsewhere we present a multivariate analysis (Sastry, in press) that examines the association between demographic, social, economic factors, and fieldwork outcomes for each case. Results for DNORPS are summarized in Table 1. The top panel shows the disposition of all cases released to the field, and the bottom panel shows several key outcome rates based on reporting guidelines from the American Association for Public Opinion Research (2006). The fieldwork results are presented for each of the three strata separately as well as for the entire sample.
In all, we successfully located 231 cases, or about two-thirds of all eligible cases. We completed 147 interviews. Of the 344 cases released to the field, 19 were ineligible. Reasons for ineligibility included having no surviving household members, a vacant dwelling at the time of Hurricane Katrina, or being unable to classify a case as a pre-Katrina household from New Orleans. Fieldwork staff located sixty-five cases that were not interviewed for various reasons. These cases were approximately evenly divided among three groups—refusals, unable-to-contact cases, and cases on which work had stopped. Unable-to-contact cases were finalized cases with an identified respondent and a current telephone number or address, but for which interviewers were unable to reach after numerous attempts. Additional attempts to contact these cases were judged to be unwarranted. Cases on which work had stopped include those with an address outside of New Orleans but no telephone number (and which hence required an in-person visit).
Not being able to locate a case (113 instances) was the main reason for not completing an interview. Some of the unlocated cases may have been ineligible, a fact that can only be determined by locating and questioning a respondent or informant. For virtually all of these cases, we had open leads that could have been pursued by the fieldwork staff but budget constraints prevented it.
The bottom panel of Table 2 presents outcome rates for the pilot study. Overall, the DNORPS field staff located 65 percent of sampled cases. In the no-flood strata, we located 76 percent of cases, declining to 62 percent in the low-flood strata and 58 percent in the high-flood strata. It was clearly more difficult to locate respondents in areas that had flooded, primarily because a higher fraction of these respondents no longer resided in the dwellings. Differences in rates of locating cases across flood strata are statistically significant on the basis of an F-test. The overall contact rate for DNORPS, calculated as the percentage of eligible located cases that were successfully contacted, was 79 percent. There was little variation across flood strata in contact rates, which ranged from 77 percent in the no-flood strata to 81 percent in the low-flood strata to 79 percent in the high-flood strata. Differences in the contact rates across strata are not statistically significant.
The DNORPS cooperation rate, calculated as the percentage of contacted cases that were interviewed, was 88 percent overall. This high cooperation rate corresponds directly to a low refusal rate of 12 percent, which was achieved even in the absence of extensive refusal-conversion efforts. Cooperation rates did not vary greatly by strata; refusal rates were 12 percent in the no-flood strata, 8 percent rate in the low-flood strata, and 15 percent rate in the high-flood strata. Differences across flood strata in the cooperation rates were not statistically significant.
At the bottom of Table 2 we present two separate response rates—an observed response rate and an adjusted response rate. In both cases, these rates reflect the percentage of eligible cases that were successfully interviewed. The adjusted rate controls for the final fieldwork stage in which we selected a subsample of cases for additional tracking and interview attempts while work stopped on the remaining cases. The adjusted response rate varies substantially by strata, from a high of 63 percent in the no-flood strata to a low of 41 percent in the high-flood strata. The overall observed response rate for the study was 45 percent, with the response rate in the no-flood strata exceeding this rate by six percentage points (51 percent) and the response rate in the high-flood strata falling six percentage points below this rate (39 percent). However, differences in the observed response rates across strata were not statistically significant.
The area-based sample design of the DNORPS allowed us to conduct a multivariate logistic regression analysis of the fieldwork outcomes, with covariates based on area-characteristics at the block and block-group level from Census 2000 (details are reported in Sastry, in press). The dependent variables are fieldwork outcomes coded at the individual level that indicate whether an eligible case was located, a located case was contacted, or a contacted case cooperated (that is, was interviewed). In addition, we examined the individual-level analog of the overall response rate by modeling whether an eligible case was successfully interviewed. Covariates in the model included sample-design variables (flood-depth strata and whether the case was named or unnamed), and census-based areal measures of demographic characteristics (age and percentage of population that was Black), socioeconomic status (percentage of employed adults with executive or professional occupations and median family income), residential stability, and percentage of vacant dwellings.
The regression results revealed few systematic differences in fieldwork outcomes by demographic, socioeconomic, and housing characteristics, across any of the fieldwork stages—with the exception of locating sampled cases. We found that cases classified as “older” (based on the average age of residents in the New Orleans block of the dwelling) were located at higher rates. This may be because households in neighborhoods with older residents are more established and have stronger ties to the neighbors and to the local area. Cases in tracts with a higher fraction of non-family households had lower rates of being located. This reflects the greater likelihood of respondents living alone or with roommates and correspondingly higher levels of transiency and weaker ties to the dwelling, the neighborhood, and the city. Finally, none of the covariates describing race or socioeconomic status were statistically significant. There were no notable effects of covariates on rates for contacting or completing cases or for the overall response rate model; the only statistically significant result was the lower probability of a cases completing the survey in the high flood area.
The 147 households successfully interviewed reported information on a total of 386 individuals who were residing in the sampled households prior to Hurricane Katrina, for an average of 2.6 individuals per household. Table 2 lists the demographic, socioeconomic, and housing characteristics of these individuals using DNORPS sample weights.
The sample is divided fairly uniformly across four broad age groups, with 27 percent under age eighteen; 19 percent aged eighteen to twenty-nine; 29 percent between ages thirty and forty-nine; and 24 percent age fifty and older. Slightly more than one-half of the individuals (53 percent) are female. Four-fifths were born in Louisiana, reflecting minimal in-migration and the slow, steady population decline in New Orleans prior to Hurricane Katrina (Brookings Institution 2005). Approximately two-thirds (67 percent) of individuals in the DNORPS sample are black and 28 percent are white. These results align closely with those from the 2004 American Community Survey, which estimated that 69 percent of the New Orleans population was black and 28 percent white (U.S. Census Bureau 2005).
Marriage, education, and employment rates are shown only for individuals age eighteen and older. Two-fifths of the adult sample is married, while a similar fraction has never been married. One-fifth of the sample is separated, divorced, or widowed. A plurality of adults (49 percent) are high school graduates, 37 percent are college graduates, and 14 percent have no high school degree. The majority of the adults (57 percent) are employed, with 15 percent unemployed and 15 percent retired.
Only 5 percent of individuals in the sample had pre-hurricane residences that had not been damaged by wind or flooding, even though 25 percent of respondents had lived in the non-flooded strata. About one-quarter of individuals had damaged but habitable pre-hurricane homes, while 71 percent of respondents’ pre-hurricane homes were uninhabitable (49 percent uninhabitable and 22 percent destroyed). One-third of the individuals had lived in a rented dwelling in New Orleans prior to Katrina and two-thirds had lived in a family-owned home. The most common destinations for displaced individuals were Texas (44 percent) and elsewhere in Louisiana (43 percent).
One year after the hurricane struck, 49 percent of pre-Katrina residents in our sample had returned to New Orleans (Table 3). The most widely accepted pre-Katrina estimate of the population of New Orleans was 454,863 in July 2005 (U.S. Census Bureau, 2005). Therefore, the population of New Orleans in fall 2006 most likely included approximately 222,900 returned residents (plus any new residents who had not resided in the city prior to the hurricane). This estimate is very similar to other independent estimates, including those from the U.S. Census Bureau (2007), which estimated the July 1, 2006, population to be 223,000, the Louisiana Health and Population Survey, which estimated the city’s total population in fall 2006 at approximately 200,000 (LPHI 2007), and the Kaiser Post-Katrina Baseline Survey, which placed the population in fall 2006 at approximately 221,000 (Kaiser Family Foundation 2007).
Table 3 presents the estimated percentages of pre-Katrina New Orleans residents who have returned to the city by individuals’ demographic, socioeconomic, and housing characteristics. All of the percentages are weighted, and we examine differences in return rates by individual characteristics using a sample design-based F-test for which we also report the statistical significance. We find large and statistically significant disparities in return rates by level of housing damage and employment and, to a lesser extent, by age, race, education, and flooding strata. We find marginally significant differences in return rates by sex and state of birth but no differences by marital status, housing tenure (owned versus rented), or displaced location. Of those who had still not returned to the city by the time of the survey, 37 percent lived elsewhere in Louisiana, 34 percent lived in Texas, 9 percent in Georgia, and 20 percent lived in various other states (results not shown).
Disparities in return rates are most pronounced by housing damage. Twice as many individuals whose homes were undamaged (86 percent) or damaged (83 percent) had returned as those whose homes were uninhabitable (Table 3). Those with undamaged or damaged homes were four times as likely to return as those whose homes were destroyed. Differences in return rates among these groups are statistically significant at the .01 level. Flooding is closely related to housing damage, but we find much smaller differences in return rates by flooding strata than by housing damage. In the unflooded strata, nearly three-quarters of pre-Katrina residents had returned, while in the low-flood strata, about one-half had returned. In the high-flood strata, 38 percent of pre-Katrina residents had returned. Return rate differences by flooding strata are statistically significant at the .05 level.
Employment also affects return rates. Individuals who are employed or retired have high return rates (63 percent and 59 percent, respectively) (Table 3). Those who are unemployed, students, or homemakers have return rates that are only one-half those of the first group (27 percent to 29 percent). Differences in return rates are statistically significant at the .01 level. Those with college degrees are more likely to have returned (69 percent) than those who have a high-school degree or less (return rates of 45 percent to 47 percent). Older adults (aged thirty and older) are also more likely to return; 58 percent to 68 percent have returned compared with 38 to 39 percent of children and young adults under age thirty. Finally, 40 percent of blacks had returned while 69 percent of whites had returned. Differences in return rates by education, age, and race are all statistically significant at the .05 level.
These results portray a consistent picture of pre-Katrina residents of New Orleans who have returned and who remain displaced. Individuals who are older, white, college-educated, employed or retired, with a habitable home were more likely to return to New Orleans in the first year after Hurricane Katrina struck. In contrast, those who remain displaced are largely children and young adults, blacks, the less educated, the unemployed, and those without a habitable dwelling in New Orleans. These distinctions suggest that the still displaced are likely to face significant obstacles to returning, including those related to schooling (for children), employment (for those without a job or out of the labor force), and in securing housing (for those with uninhabitable homes).
We asked all respondents to report, on a scale of zero to ten, the likelihood of living in New Orleans in one year’s time (that is, fall 2007) for each resident in the respondent’s pre-Katrina household. A zero indicated no chance that the person would be living in New Orleans and a ten meant that it was absolutely certain that the person would be living in the city. We created a three-category variable from the response scale, with zero to two indicating that the person was not likely to return to New Orleans, three to seven indicating the person might return, and eight to ten indicating the person was likely to return. We chose these categories after examining detailed tabulations of the data. The results, presented in Table 4, are not sensitive to the use of different categories.
Overall, two-thirds of pre-Katrina residents, regardless of where they lived at the time of the interview, were possibly (18 percent) or very likely (48 percent) to be residing in the city in fall 2007. We can use this figure to estimate the population of New Orleans in fall 2007. We recognize, however, that this estimate will exclude new residents of the city and will include people who report that their likelihood of returning is uncertain—and that there is no way to know the relative sizes of these two groups. Nevertheless, the two-thirds figure is consistent with estimates from other sources. In particular, the Brookings Institution and the Greater New Orleans Community Data Center (2008) estimated that in late 2007 the population of New Orleans was about 70 percent of its pre-Katrina size based on U.S. Postal Service counts of households receiving mail. The Brookings/GNOCDC estimate includes new residents of New Orleans but does not identify the fraction of the population composed of new entrants and of displaced residents who have returned.
The second and third panels of Table 4 show the dramatic differences in the likelihood of living in New Orleans in one year’s time for residents living in the city and for residents living elsewhere. Overall, among pre-Katrina residents living in New Orleans, 97 percent report being possibly (8 percent) or very (89 percent) likely to be living in the city in one year’s time. This share varies little by flood strata. Very few residents (3 percent) planned to leave the city in the coming year. On the other hand, only 13 percent of pre-Katrina New Orleanians residing elsewhere said they were very likely to return to the city in the coming year, while 26 percent said they were possibly going to return. Here we see substantial differences by flood depth. Only about one in five of displaced residents from the unflooded strata said they were likely to return to the city in the coming year. In other words, those who could have returned but chose not to are unlikely to change their minds and return to the city. In contrast, nearly one-half of residents from the low-flood zone and four in ten from the high-flood zone said they were likely to return in the coming year.
In examining the characteristics most likely to shape returns to New Orleans among displaced residents we find only three that are related the likelihood of return. The first and largest disparity is by place of birth—displaced individuals who were born in Louisiana (and quite possibly in New Orleans itself) have much higher expectations of returning to the city than those displaced individuals who were born elsewhere (Table 5). In addition, those who had lived in rented housing prior to Hurricane Katrina were more likely to return than those who had owned their home. Most likely, homeowners who wanted to return already had, while renters who desired to return were still unable to. It is possible that many renters have unreasonably optimistic expectations about being able to find housing in New Orleans. Finally, there is a puzzling disparity by marital status, with married individuals far less likely than the never-married or separated or divorced or widowed individuals expecting to return to the city. Overall, these findings suggest that among the New Orleans residents who remain displaced, those with the strongest desire to return may be the least able to do so because they lack housing, family, and perhaps other necessary resources.
The main goals of Displaced New Orleans Residents Pilot Study were to establish the feasibility of identifying a representative sample of pre-Katrina residents of New Orleans and to determine the likely success of efforts to track and interview this sample. An additional aim was to use the pilot data to gain insights into the recovery (as measured by its future population) of New Orleans in the aftermath of Hurricane Katrina and to study the well-being of displaced residents.
The DNORPS results showed that sampled respondents once contacted could be successfully interviewed, with cooperation rates approaching 90 percent, but that the main challenge was in locating respondents. Also, there were few systematic differences in fieldwork outcomes by demographic, socioeconomic, and housing characteristics across any of the fieldwork stages—with the exception of locating sampled cases. The main challenge in improving the fieldwork results and in launching a full-scale study of this population is to locate a high fraction of sampled cases.
In all, we estimate that the New Orleans population will regain over two-thirds of its residents who were displaced by Hurricane Katrina. At the time of the survey in the fall of 2006 approximately one-half of residents had returned to New Orleans and nearly all of those planned to stay. We found essentially no variation in this high likelihood of remaining in the city by flood zone, which is remarkable given the vastly different living environments of flooded and non-flooded residents. At the same time, among the 51 percent of residents who had not yet returned, the likelihood of moving back is modest. Of those displaced, the majority (61 percent) had no plans to return within a year of the interview (i.e., within two years following the hurricane). Interestingly, the likelihood of returning to the city is higher among residents in flooded parts of the city, which likely reflects the effects of unfinished reconstruction efforts, such as repairing or replacing damaged housing and waiting for neighborhood infrastructure to be reestablished. In unflooded parts of the city, few additional displaced residents are likely to return if they have not done so already. The overall implication of these results for the future population of New Orleans is likely only very modest growth from the return of still-displaced residents. The city will continue its post-Katrina experience of being older, whiter and more highly-educated, and with fewer families, children, and people out of the labor force. Our results suggest that perhaps the most effective way to increase the proportion of displaced New Orleanians who return to the city is to increase the availability of low-cost housing to renters who are still away from the city but would like to return.
The author gratefully acknowledges the contributions of many colleagues in designing, implementing, and analyzing the Displaced New Orleans Residents Study, including Megan Beckett, Sandro Galea, Katherine Hall, Amelia Haviland, Anne Kenyon, Karol Krotki, Nicole Lurie, Adrian Overton, Kavita Patel, Christine Peterson, Michael Rendall, Ben Springgate, Greg Stone, Susan Twiddy, Joseph McMichael, Lily Trofimovich, Mark VanLandingham, Clayton Williams, and Julie Zissimopolous.