Katrina and Migration: Evacuation and Return by African Americans and Vietnamese Americans in an Eastern New Orleans Suburb

Abstract

Hurricane Katrina constitutes the most costly natural as well as technology-induced disaster, in terms of both human suffering and financial loss in the history of the United States. Even years later, it continues to profoundly impact the livelihoods and the mental and physical health of those who have experienced evacuation and return and those who have begun lives anew elsewhere. Our study focuses on these geographical processes associated with the Katrina disaster experiences of African Americans and Vietnamese Americans comprising an overwhelming majority (93.4 percent) of residents in a racially mixed pre-Katrina eastern New Orleans neighborhood. We examine the spatial morphology of routes, volumes, and frequencies of evacuees; their return rates and experiences; and rationales and motivations to return or stay. The conceptual framework is based on the disaster migration, place attachment, and social network literature. Both quantitative and qualitative evidence indicates that the evacuation and return experiences of each minority group substantially differed, especially among African American women, and this was strongly influenced by existing social networks.

El huracán Katrina es el desastre natural, también inducido por la tecnología, más costoso en términos de sufrimiento humano y pérdidas económicas en la historia de los Estados Unidos. Incluso años después de ocurrido, continúa impactando profundamente las vidas y salud física y mental de quienes experimentaron la evacuación y el regreso, y de quienes decidieron empezar todo de nuevo en otra parte. Nuestro estudio se concentra en estos procesos geográficos asociados con las experiencias del desastre del Katrina entre los afroamericanos y americanos de origen vietnamita que comprenden una abrumadora mayoría (93.4 por ciento) de los residentes de diferentes razas de un vecindario pre-Katrina en el oriente de Nueva Orleans. Examinamos la morfología espacial de las rutas, volúmenes y frecuencias de los evacuados; sus tasas de regreso y sus experiencias; y las razones y motivaciones que tuvieron para regresar o permanecer. El marco conceptual se basa en la migración por desastre, apego al lugar y literatura de redes sociales. Tanto la evidencia cuantitativa como la cualitativa indican que las experiencias de evacuación y regreso de cada grupo minoritario difieren sustancialmente, especialmente entre las mujeres afroamericanas, y eso estuvo fuertemente influido por las redes sociales existentes.

Key Words:

• African American
• evacuation and return
• Katrina
• New Orleans
• Vietnamese American

关键词:

• 非裔美国人
• 撤离和返回
• 卡特里娜飓风
• 新奥尔良
• 越南裔美国人

Palabras clave:

• afroamericano
• evacuación y retorno
• Katrina
• Nueva Orleans
• vietnamita americano

Hurricane Katrina constitutes the most costly natural as well as technology-induced disaster in terms of both human suffering and financial loss in the history of the United States. Even years later, the storm's aftermath remains a topic of significance for the news media, American politics, and the international image of the United States (Archibald and Munn-Venn 2007). More important, it continues to profoundly impact the livelihood and the mental and physical health of those who experienced evacuation (Jones and McDaniel 2006) and returned to and rebuilt in New Orleans and those who have begun lives anew elsewhere. The media's coverage of continued debates about how to best rebuild New Orleans, the struggles of homeowners with insurance claims, the lack of housing available to displaced renters (many of whom are ethnic minorities), and the lack of attention to specific neighborhoods and schools reveal the degree of racial inequality and economic disparities that continue to blight American society. The scope of Katrina evacuees' continued struggles to make a living and resume a normal life, either in New Orleans or elsewhere, is unprecedented in recent memory.

Academic studies about Katrina have multiplied, as journal special issues and books devoted to Katrina attest. This research ranges from contextual analyses—examining racial and class disparities, mental and physical health impacts—to political implications of Katrina (Bobo and Dawson 2006; Leong et al. 2007). Publications either mourned the loss of New Orleans as a site of American culture or used the government's inadequate response to Katrina as a platform for further theorizing about advanced capitalism's legacy of class and racial divisions and the dehumanization of the marginalized in the United States (Brinkley 2006; Dyson 2006; Piazza 2006). However, none of the existing research addresses the issue of decision making with reference to the spatial morphology of evacuation (how many moves and where) or return, which remains a gap in the disaster literature.

Additionally, academic studies of Katrina and other disasters seldom addressed the impacts on Asian Americans (Kao 2006), partially because the total population of Asian Americans was too small to obtain an adequate survey sample size. Even the special data product on the impacts of Hurricanes Katrina and Rita released by the U.S. Census Bureau (2005), “2005 ACS Special product for the Gulf Coast Area,” failed to mention Asian American victims. This population was also neglected by the mainstream news media in the immediate aftermath of the storm. This neglect, however, was followed by a flurry of coverage of Vietnamese Americans in New Orleans East, perpetuating the “model minority” myth that Asian Americans handled their post-Katrina community affairs internally rather than relying on external assistance (Leong et al. 2007). The other important issue is that most existing Katrina research focuses on a racial binary of white and black relations and experiences (Cutter and Emrich 2006), with little attention paid to minority–minority comparisons.

Therefore, the need remains to critically and comparatively examine the evacuation and return geography of Katrina evacuees, especially among those who are socially vulnerable to natural and technology-induced disasters, and to answer the following questions:

1.	What were the evacuation routes, volumes, and frequencies exhibited by evacuees? What explains these evacuation patterns?
2.	What were evacuee return rates and experiences and rationales and motivations behind the decision to return or stay?
3.	What primary factors are impacting evacuation and return? How important are the roles of place attachment and social networks?

To address these questions, we assembled a team of multidisciplinary scholars to conduct a research project that is interdisciplinary in nature and uses a mixed-method approach: field observations, formal surveys, focus group discussions, in-depth interviews, and mapping. The project studies one racially mixed eastern New Orleans neighborhood, where African Americans and Vietnamese Americans combined comprise 93.4 percent of pre-Katrina residents. It aims to address gaps in the literature by comparing two racial minority groups and presenting results from a survey complemented by personal narratives of Katrina experiences. This article presents one aspect of our research results on evacuation and return geography in which we first offer a conceptual framework of disaster migration, place attachment, and social networks among vulnerable people, followed by a methods discussion and a demographic description of the study area. We then analyze the evacuation geographies of survey participants and examine their return process.

Disaster Migration, Place Attachment, and Social Networks

Unlike economic migrants, disaster migrants often consider their relocation temporary and desire to return after the immediate destruction and damage are over. Katrina, however, has resulted in unique scenarios and consequences wherein a high percentage of evacuees have decided to remain at their evacuation destinations. As revealed by U.S. Census figures depicting metropolitan population changes between July 2005 and 2006, New Orleans experienced a dramatic decline in population that was mirrored by strong growth in urban areas such as Houston, Texas, and Baton Rouge, Louisiana. Indeed, the number of households in New Orleans in June 2008 was just 71.8 percent of its pre-Katrina June 2005 total (Greater New Orleans Community Data Center 2008).

Recent research on the demography of disasters has addressed the impact of disasters on marginalized communities, their preparedness, and response. These studies demonstrate that racial and ethnic communities face more obstacles to disaster preparedness and as a result experience greater anxiety about potential disasters and are less likely to respond to disaster warnings. Pastor et al. (2006) suggested that delayed evacuation is common among minority homeowners because home ownership disproportionately represents the wealth of ethnic minorities and those who are less affluent, which makes it hard to leave their houses behind.

Hunter (2005) noted in a comparative overview of global natural disasters that migration decisions related to disaster vary by hazard type, settings, household characteristics, and perception of risk, and that this variance is particularly visible among vulnerable populations. Human assessment of risk may be shaped by social factors specific to place, such as place attachment, or economic limitations on geographic mobility. One adjustment to a hazard is migration or redistribution of the population. Hunter cited Hugo's continuum of mobility from migration based on voluntary choice to that entirely forced by threat of death and applied this to a community model of migration in response to environmental hazards: A community may reconstruct its prehazard state, partially reorganize (e.g., spatially) to reduce risk, or, most rarely, completely relocate to avoid risk altogether. In addition, households might themselves employ different strategies, wherein certain members move permanently as others remain, sometimes resulting in the entire household relocating in response to the negative economic impact of the disaster. Hunter cited Morrow-Jones and Morrow-Jones's 1991 study, which indicated that migrants who leave as a result of a natural disaster “are more likely female-headed households and minority group members, and are characterized by lower income and educational levels” (Hunter 2005, 286).

Based on literature concerning the “psychology of place” in geography, psychology, anthropology, and psychiatry, Fullilove (1996, 1516) has identified some key processes: “place attachment… is a mutual caretaking bond between a person and a beloved place.… Place identity is concerned with the extraction of a sense of self based on the places in which one passes one's life.” After a major life interruption event, such processes are “threatened by displacement, and the problems of nostalgia, disorientation, and alienation may ensue.” This is especially true following environmental disasters (Brown and Perkins 1992). Recognizing that attachment to place varies by socioeconomic and demographic characteristics, the stronger the predisaster place attachment, the more severe the suffering experienced after forced displacement. Place attachment to one's initial neighborhood would, in turn, impact the decision to relocate or return, as well as one's overall social resiliency during and after catastrophes. Gerrity and Steinglass (1994) revealed the meanings to families of losing a home after a disaster and the resulting coping strategies. Similarly, Dugan (2007) illustrated, via the experiences of a Katrina victim who permanently settled in Texas, that evacuees are able to adopt another cultural identity over time in their new home. However, what they lost in New Orleans—their homes, cultural milieu, and identity—cannot be easily replaced.

One of the many critical mitigating factors that impact the experiences of disaster evacuees is the presence or absence of social networks. Conceptually, social networks are comprised of “the trust, mutual understanding, and shared values and behaviors that bind the members of human networks and communities and make cooperative action possible” (Cohen and Prusak 2001, 4). Indeed, attachment to place and community is dependent on social capital embedded in social networks (Bridge 2002). With reference to social networks and disaster research, social networks “facilitate a flow of information providing a basis for action and assisting in individual and community goal attainment” (Ritchie and Gill 2007, 109). Especially in the context of disasters, religious institutions are a foci of social networks because “[I]nstitutional structures can make a difference to levels of participation and thereby, influence the formation of social capital” (Mohan and Mohan 2002, 197). Thus, ethnic-based religious institutions have always been a source of ontological trust, especially compared to larger scale institutions such as governments (Fennema and Tillie 2001).

The important intersecting roles of social networks and evacuation and return experiences in the context of hurricane-induced disasters are supported by existing research. In studying Hurricane Andrew's impacts on African American elderly who lived in a public housing complex, Sanders, Bowie, and Bowie (2003) concluded that the majority of elderly women in economically depressed low-income communities suffered tremendous physical and mental health problems as a result of being separated from their normal social support system. The National Congress of Vietnamese Americans report on housing for Vietnamese communities in Biloxi and New Orleans affected by Katrina noted that the presence of social networks (family, friends, religious institutions) was a factor for those who chose to return, even as the smaller numbers of those who decided to not return also cited the dispersal of Vietnamese Americans, their families and friends moving away, in addition to a lack of open public schools for their children, as their top reasons to leave (National Congress of Vietnamese Americans 2006). These reasons support research that attributes relocation choices to social networks that help mitigate the risks of moving—or not moving, as in the case of New Orleans post-Katrina (Massey 1990; Pastor et al. 2006). McCarthy et al. (2006, 9–10) noted that the impacts of social networks are cumulative because “decisions are likely to be strongly shaped by what friends, co-workers, family members, and neighbors decide to do.” With reference to the important role of religious institutions in disaster mitigation, in the few days following the 1992 landfall of Hurricane Andrew near Miami, churches were the only institutions providing immediate relief to various neighborhoods (Morrow and Peacock 1997).

Although research addressing the various human impacts of Katrina is now plentiful, empirical analyses on Katrina evacuees are limited. Part of this is due to, as Smith and McCarty (1996, 265) concluded in their case study of the demographics of Hurricane Andrew evacuees and returnees in 1994 (with a dispersal of 333,000 persons, and much smaller number of permanent evacuees at 40,000), “the scarcity of timely, accurate, and comprehensive data.” Based on a Gallup poll among a representative sample of 1,294 African American and white evacuees a month after Katrina, Elliott and Pais (2006) revealed the intricacies of race and class in Katrina victims' evacuation time, short-term recovery, postdisaster coping, and likelihood of return. In particular, they concluded that lower income African Americans were more likely to remain in New Orleans during the disaster, and that home ownership and household income, more than racial differences, influence people's desire for return, although lower income homeowners were more likely to return, largely due to lack of financial options.

Study Area and Research Methods

Study Area

Our study area is one of the many distinctive New Orleans communities and is portrayed as a Vietnamese enclave by the media despite its multiracial nature (Airriess 2002). Separated from the rest of the city by I-510, it is a geographically isolated community made up of two of the most easternmost suburban census tracts (17.41 and 17.42) in the New Orleans Metropolitan Statistical Area (MSA) that were carved out of marshland and swamp beginning in the 1960s (Figure 1). In 2000, African Americans (51.7 percent) and Vietnamese (41.7 percent) comprised a majority of the study area's population. Pre-Katrina New Orleans East as a whole, and the study area in particular, included a large African American population of varied socioeconomic status, despite being misperceived as a Black ghetto (Johnson 2003). Compared to the New Orleans MSA, our study area possessed higher percentages of renter and poverty households among African Americans and Vietnamese Americans and higher unemployment rates for African Americans. The adult population was less educated and the median household income was lower for both groups. African Americans had a higher percentage of females and a lower percentage of private vehicle ownership, whereas Vietnamese Americans had much higher percentages of foreign-born and linguistically isolated households (Table 1).

Figure 1 Study area.

Table 1 Demographic and socioeconomic characteristics

			Study area: 2000		Survey respondents
	Damaged area total	New Orleans Metropolitan Statistical Area: 2000	African American	Vietnamese American	African American	Vietnamese American
Total population	> 560,000	1,337,726	5,450	4,461	40	99
African Americans (%)	45.8	37.5	51.7	41.7	28.8	71.2
Unemployment (%)	7.6	4.1	10.5	3.1	N/A	N/A
Renter households (%)	45.7	39.0	59.7	47.8	30.6	41.1
Poverty households (%)	20.9	18.3	34.6	31.3	N/A	N/A
1999 median household income	N/A	$35,317	$20,753	$32,001	N/A	N/A
25 years or older without bachelor's degree (%)	N/A	78.5	88.0	87.6	70.0	79.8
Female (%)	54.0	52.1	56.3	48.1	60.0	46.5
Foreign-born (%)	N/A	4.7	1.6	32.5	2.5	75.5
Linguistically isolated households (%)	N/A	1.8	0.3	36.7	N/A	N/A
Vehicle ownership (%)	N/A	85.0	68.0	83.0	N/A	N/A

The 2000 Census informs us of the differing socioeconomic characteristics of African Americans and Vietnamese Americans in the study area. Vietnamese Americans have larger but fewer female-headed households compared to those of African Americans, with predominantly family members and relatives making up households in both groups. The average size of African American households was 3.9 members, compared to 4.7 members for Vietnamese American households. Households with seven or more members, however, differed substantially between these two groups: For African Americans the rate was 4.5 percent, but for Vietnamese Americans it was 15.2 percent. There also existed a substantial difference in the gender of the head of household. For African Americans, 57.7 percent of households were headed by females compared to only 10.5 percent for Vietnamese Americans. With reference to the elderly population, 3.8 percent of African Americans were sixty-five years or older compared to 8.0 percent for Vietnamese Americans. Households were overwhelmingly made up of relatives; nonrelatives made up 2.4 percent of households for African Americans and only 1.3 percent for Vietnamese Americans.

African Americans and Vietnamese Americans were unevenly distributed throughout the study area. Prior to Katrina, residential spaces consisted of single-family homes, duplexes, and three apartment complexes, including a Section 8 complex with 186 units. African Americans were a residential minority in forty of the fifty-nine census blocks east of Michoud Boulevard. The three apartment complexes, however, are also located in this section and housed 38.6 percent of the area's pre-Katrina African American population, compared to only 6.6 percent of Vietnamese Americans. The neighborhoods west of Michoud Boulevard featured an African American majority, with less than 30 percent of its residents Vietnamese Americans.

The Vietnamese community is extremely tight knit in part because over half of the original population who arrived as “first wave” refugees in 1975 originated from a handful of villages in North Vietnam (Airriess 2002). In a chain migration process, relatives and friends who were scattered across the United States relocated to the New Orleans community during the 1980s and 1990s. The close-knit nature of the community is also based on a common faith because approximately 80 percent of residents are Roman Catholic (Nash 1992; Airriess 2002), and this faith is anchored by the large Mary Queen of Vietnam Church. Indeed, this parish church is the single large public institution in the community. This community represents the most dense settlement concentration of Vietnamese Americans in the country. In New Orleans, the community accounted for approximately 30 percent of the 14,868 Vietnamese Americans in the New Orleans MSA in 2000, whereas African Americans in our study area accounted for only 1 percent of the entire New Orleans MSA African American population during the same year. An even greater contrast is that Vietnamese Americans owned all ninty-nine businesses in the area prior to Katrina, with no African Americans owning businesses (Airriess 2006). The African American population, especially middle-class homeowners, are far more heterogeneous, having been part of the urban exodus into suburban New Orleans East beginning during the 1960s. No single religious institution is the center of spiritual and social life because churches are primarily located outside the physical community and are many in number.

Research Methods

The comparative results presented here are primarily based on a survey of forty African Americans and ninety-nine Vietnamese Americans from our study area from December 2005 to December 2006, all of whom lived in our study area prior to Katrina. ¹ Results are also drawn from two focus group discussions within each group conducted in March and June 2006, respectively, and each lasting at least one hour; in-depth interviews with community and religious leaders; plus field observations and conversations during eight trips to New Orleans or Houston.

Similar to Smith and McCarty (1996) and Elliott, Sams-Abiodun, and Haney (2007), we faced tremendous survey administration difficulties in reaching potential respondents under the postdisaster context. The “Katrina Diaspora” phenomenon (Ericson, Tse, and Wilgoren 2005) made it challenging to reach pre-Katrina African American renters from our study area in particular, given that none of the three apartment complexes were rebuilt and a majority of the former renters never returned during our one-year survey period. To make collected information comparable, as we ask many time-sensitive questions, we had to stop data collection at the end of the survey period. This, unfortunately, caused large cross-group sample size disparity.

Table 1 compares selected demographic and socioeconomic characteristics of the pre-Katrina New Orleans MSA and the overall Katrina-damaged Gulf Coast area with those of our survey respondents. To recruit our survey respondents, we used purposive sampling, a potentially effective method for research that aims to understand and obtain information from a target population and where sample representativeness might not be the primary concern (Trochim 2006). Thus, our sample is not statistically representative of the population in the study area for the aforementioned reasons (with the exception of similar percentages of females), and certainly not representative of the New Orleans MSA as a whole, but serves our research purpose. Our sample overrepresents Vietnamese Americans, homeowners, people with higher education, and those who are foreign-born. Our data biases also include a large number of early returning Vietnamese American respondents participating in a major local community event in February 2006. African American participants are primarily returned homeowners who were administered surveys after four months as evacuees.

Relevant parts of the survey for this article include demographic and socioeconomic characteristics, community ties and place attachment prior to and post-Katrina, evacuation routes and frequencies, and reasons for return. Focus group discussions included questions about participants' disaster migration histories, place attachment, Katrina experiences, and the role of community leadership, as well as language barriers among the Vietnamese-speaking-only population in obtaining information and seeking assistance.

Evacuation Geography

The chaos caused by Katrina severely impacted evacuees' lives regardless of when, how, or where they evacuated. However, such experiences did vary by race, class, and gender. Table 2 presents survey summary statistics on evacuation, return, and reasons associated with both processes. The 139 survey respondents from our study area evacuated to a total of thirty-seven different cities in thirteen states (Figure 2). This section describes their evacuation experiences.

Figure 2 (A) First, second, and third evacuation destinations, African Americans. (B) First, second, and third evacuation destinations, Vietnamese Americans.

Table 2 Summary survey results

	African Americans		Vietnamese Americans
	Male ^a	Female ^b	Male ^c	Female ^d
% evacuated prior to Katrina	85.7%	84.6%	96.2%	93.5%
Evacuation mode
Car	91.7%	86.4%	98.0%	95.3%
Bus	0.0%	4.5%	0.0%	0.0%
Other transportation modes (including boat)	8.3%	9.1%	2.0%	4.7%
Evacuation statistics
Average number of evacuation stops	2.07	2.38	1.45	1.41
Evacuation frequencies
% no evacuation (stayed in New Orleans)	7.1%	7.7%	7.5%	13.0%
% experienced one evacuation	21.4%	19.2%	60.4%	58.7%
% experienced two evacuations	50.0%	30.8%	18.9%	15.2%
% experienced three evacuations	7.1%	23.1%	9.4%	4.3%
% experienced four evacuations	7.1%	7.7%	0.0%	4.3%
% experienced five evacuations	7.1%	11.5%	3.8%	4.3%
% evacuated with family
First evacuation	100.0%	85.0%	92.5%	97.1%
Second evacuation	100.0%	76.9%	92.3%	100.0%
Third evacuation	100.0%	77.8%	100.0%	100.0%
Fourth evacuation	0.0%	100.0%	100.0%	100.0%
Fifth evacuation	0.0%	100.0%	50.0%	100.0%
% having relatives at destination
First evacuation	54.5%	50.0%	80.0%	52.9%
Second evacuation	75.0%	50.0%	71.4%	77.8%
Third evacuation	50.0%	66.7%	57.1%	80.0%
Fourth evacuation	100.0%	40.0%	50.0%	66.7%
Fifth evacuation	0.0%	0.0%	50.0%	50.0%
Returning to neighborhood
% plan to return	35.7%	38.5%	45.3%	43.5%
% plan to stay in evacuated place	35.7%	26.9%	15.1%	23.9%
% returned within two months post-Katrina	7.1%	15.4%	20.8%	17.4%
% not returned after three months	28.6%	38.5%	9.4%	13.0%
Doing a good job during Katrina to get us informed, evacuated, and provide disaster assistance
My church (range = 0–5)	1.92	1.17	3.57	3.73
My community leaders	1.50	1.13	3.78	3.53
Feeling confident that we can be united in the community vision we present to outsiders prior to Katrina
None	7.7%	4.2%	6.0%	2.3%
Rarely	7.7%	20.8%	20.0%	9.3%
Sometimes	23.1%	37.5%	22.0%	44.2%
Mostly	61.5%	37.5%	52.0%	44.2%
Feeling confident that we can be united in the community vision we present to outsiders post-Katrina
None	7.1%	17.4%	3.9%	4.5%
Rarely	35.7%	34.8%	15.7%	15.9%
Sometimes	14.3%	21.7%	29.4%	27.3%
Mostly	42.9%	26.1%	51.0%	52.3%
Confident that our community can create adequate resources to develop new jobs despite changes in the economy prior to Katrina
None	21.4%	4.3%	29.4%	13.6%
Rarely	0.0%	21.7%	15.7%	2.3%
Sometimes	42.9%	52.2%	37.3%	45.5%
Mostly	35.7%	21.7%	17.6%	38.6%
Confident that our community can create adequate resources to develop new jobs despite changes in the economy post-Katrina
None	35.7%	27.3%	22.4%	11.9%
Rarely	21.4%	27.3%	14.3%	16.7%
Sometimes	28.6%	40.9%	32.7%	28.6%
Mostly	14.3%	4.5%	30.6%	42.9%
Treated badly due to racial minority status
Prior to Katrina (range = 0–5)	2.14	2.04	1.88	1.70
During Katrina	3.23	2.44	2.08	1.30
; ; ; and

Evacuation Frequencies and Routes

Our survey reveals that the overwhelming majority of both African American and Vietnamese American respondents, male and female, evacuated by car prior to Katrina's landfall. However, a higher percentage of African American evacuees, especially women, used other means to evacuate compared to Vietnamese Americans. Similarly, African Americans reported more evacuation stops (i.e., being relocated, which includes returning to New Orleans) than Vietnamese Americans. African American women in particular had the highest average number of evacuation stops (2.38) among all four groups. Vietnamese females were most likely to remain in New Orleans during Katrina (13 percent). About 60 percent of Vietnamese Americans and 20 percent of African Americans evacuated only once. African American females fared the worst, as 42.3 percent evacuated at least three times when compared to African American males (21.3 percent). Only 12 percent of Vietnamese Americans evacuated at least three times. Contrary to popular belief that women are more likely to evacuate with their families, African American women reported the highest percentage of evacuating without their families in the first three evacuations (Table 2).

Evacuation Destinations and Reasons

With few exceptions, Vietnamese Americans, both male and female, were more likely in the first and subsequent evacuations to evacuate to destinations where they had families or relatives than were African Americans (Table 2). Having existing social networks at their destination would likely provide both emotional and material support to Katrina evacuees during the critical moments of initial evacuation. Houston was the top evacuation destination for both groups. However, Houston as a destination was more important for Vietnamese Americans than African Americans among our survey respondents. Houston was the single most important destination for Vietnamese Americans during the initial evacuation (46.7 percent) when compared to nearby Baton Rouge (12 percent) and the third most important stop of Lafayette (8 percent). These three destination cities account for the initial evacuation of two thirds of all Vietnamese Americans. Houston was also the second evacuation destination for 34.8 percent of Vietnamese. The spatial pattern of African American evacuation, however, was substantially different, with a different order of the top three evacuation destinations and lower percentages. These evacuation destinations were Houston (19.2 percent), Lafayette (15.4 percent), and Baton Rouge (11.5 percent), but these top three cities accounted for fewer than half of African American evacuees.

Such different destinations between the two groups testify to the regional scale of social capital and social networks between the Vietnamese American community in New Orleans East and Houston (Airriess et al. 2008). The connections between Vietnamese in New Orleans as a whole and Houston are similarly strong. In three New Orleans ZIP codes, 50 percent of all Vietnamese Americans evacuated to Houston. When asked why they chose to evacuate to Houston, some Vietnamese American focus group participants in Houston responded, “Like you say, why do you have to move to Houston? There are more Vietnamese…” “It just basically because there are more Vietnamese here.…”

The following quote is particularly telling about the importance of Houston as a communication center among Vietnamese Americans during and after Katrina and the role of the parish church and its priests:

After I evacuated, there were several hundreds of people still there, the radio said that they evacuated already, but I didn't know if they actually did evacuate, so I called Houston radio station [Saigon Radio] that I'm looking for my relatives, including my husband and some friends. I asked them if they knew any info about Father Vien, Father Dung and Father [Luke] Nguyen… if Father Vien and Father Luke already evacuated, then my husband must have evacuated. When I called the radio station, they said that Father Vien already gone. Right after that Father Vien gave a speech on the Houston radio station.… (italics added)

On the other hand, Houston was just one of many destinations to which our African American respondents evacuated. For the second and third evacuations, Houston slipped to second place.

Returning Home

Different experiences and varying reasons influenced Katrina evacuees' decisions to return to New Orleans, stay in their evacuation destinations, or resettle in other places. Whereas many evacuees had no choice in terms of their evacuation destination, people made decisions to return or stay depending on their perceptions and realities of available information and resources in both New Orleans and elsewhere. Based on survey results and focus group discussions, this section specifically compares the return rates between the two racial minority groups and reveals their underlining rationales.

Returning Rates: Residential and Commercial

Vietnamese Americans in the study area returned earlier and in greater numbers when compared to the New Orleans East population in general and African Americans in the study area. Table 2 illustrates that Vietnamese Americans, both male and female, intended to return to the study area at a higher rate than that of African Americans. In contrast, African Americans, especially male, intended to remain at their evacuation destinations. In terms of actual return to the neighborhood, more Vietnamese American males and females returned within two months after Katrina compared to their African American counterparts. The contrast between the two male groups is even starker; actual return among Vietnamese American males is 20.8 percent compared to 7.1 percent for African American males. However, the percentage of nonreturning African American females three months after Katrina was the highest among all four groups (38.5 percent). One of the key reasons for the early return of Vietnamese is the supportive role of the parish church that functioned as the logistical center of the rebuilding process, as we have demonstrated elsewhere (Airriess et al. 2008).

We conducted three field surveys in early May, late June, and early November 2006 to enumerate returned households and discovered different temporal return patterns for each racial group (Figure 3). By early May, seven months after evacuating, approximately 80 percent of Vietnamese households had returned in those neighborhoods closest to the parish church; by late June, return rates had increased to 90 percent for the entire study area. In predominantly African American neighborhoods where Vietnamese were a minority, return rates for both groups ranged from 45 to 74 percent in early May. By late June, return rates in all neighborhoods except one increased to between 70 and 87 percent. Almost 40 percent of the African American population in the study area had not returned because the apartment complexes they inhabited had yet to reopen. We witnessed the first post-Katrina Vietnamese business reopening in the study area in early December 2005 and also conducted door-to-door surveys in February, early May, and late December 2006 to determine the return status of the niney-nine pre-Katrina businesses. In early February, only 25 percent of these businesses had reopened, but by early May this had increased to 59 percent, and to 90 percent by late December 2006.

Figure 3 Residential return rate, May and June 2006.

Reasons to Return

This section examines why people chose to return to their neighborhoods based primarily on place attachment and related issues of community resiliency. Although our survey and focus group discussions revealed an intimate place attachment to New Orleans as well as concerns about post-Katrina employment opportunities, we focus here specifically on local-scale place attachment.

Our survey included a number of questions about respondents' perceptions of their respective communities and the results are summarized in Table 2. Vietnamese Americans consistently gave much higher marks to their churches and community leaders for effectively dealing with Katrina compared to African Americans. This is a reflection of the close-knit nature of the Vietnamese American community. There was little in the way of racial and gender differences in assessments of the confidence individuals had about a unified community vision prior to Katrina (those who answered “sometimes” and “mostly”). However, the post-Katrina confidence level was starkly different. African American women ranked the lowest again on this measure (47.8 percent) compared to African American men (57.2 percent) and their counterparts among Vietnamese Americans (79.6 percent and 80.4 percent, respectively). Such different perceptions of their own post-Katrina community might have contributed to the differential return rates between the two communities, which in turn were functionally linked to the different structure of social networks characterizing each community.

Focus group participants from both communities exhibited obvious place attachment to their neighborhoods. Vietnamese Americans clearly viewed their community as where they wanted to be, even among those who stayed in Houston for more than six months after Katrina. The following responses speak to place attachment in different ways, but all address the importance of place as a site of social networks and belonging. For example, the following Vietnamese individual discussed the place attachment of the elderly:

The weather should be warm, so old people they like to live in a warm place, and a lot of friends, old friends, so that they can talk, they can walk.… Usually they live really close to each other that they can walk, house to house. They walk to their friends' house after dinner. That's why they have so many friends here, and they want to live here.

Comparing their own neighborhood with Houston, one Vietnamese respondent said, of New Orleans, “All our family live in the same ZIP code.… Yeah, they have a new living. Because they used to have lots of friends in the neighborhood. With old people, old people are easy to understand each other.” The shared refugee experience perhaps created a connection for the “old people,” but the younger generations were faced with developing new ties in a different location with its distinct cultural identity. Perhaps reflecting what another Vietnamese respondent lost, another respondent described the comfort of the close-knit community he associated with the study area: “I know everybody in the neighborhood, they know me by name, I know what they're doing and they do know what I'm doing, we're just like a big family. I don't want to move away from my neighborhood.”

Nonetheless, the prevailing response was that most older-generation Vietnamese refugees considered the study area the only home they knew after two previous forced evacuations as refugees: first from their homeland villages in North Vietnam to South Vietnam in 1954, then from South Vietnam to the United States after the 1975 fall of Saigon. The following respondent conflated ethnic and national identity, perhaps a reflection of how migration choices often take social networks into account:

There are many Vietnamese Americans here [New Orleans]. I like living with Vietnamese Americans. We feel like we're in our own country,… that's only the big value before Katrina, all our neighborhood, we all talk to each other, we're there just like family. Before the hurricane, I like it here because I'm Catholic and I live close to two churches. It is very convenient for both children and older people. The market is close by too, the neighborhood made me happy. I like it when there are people living around, unlike living alone. Since I came here, I always think that this is my second home, after Vietnam, from North Vietnam to South Vietnam to this place, I only want to stay here, not anywhere else. I like living here because it was fun. There are many Vietnamese Americans in Houston but I like staying here more.

Some respondents expressed their resiliency when asked to compare Katrina experiences with their lives as refugees. The following reflects how many respondents viewed the rebuilding of their community after Katrina:

In the beginning [leaving Vietnam], it is very hard because the problem is the language. The Katrina, we don't have problem with English that much, so it is easier to find a job. It is a lot different from Katrina. Our life is not as hard as before. It is harder leaving from your culture. Hurricane is nothing. In the hurricane, you have your family with you all the time.

Similarly, some African American respondents also spoke about a shared sense of place attachment prior to Katrina. A male and a female focus group member, respectively, said that African American participants who returned also acknowledged the cumulative positive impact of people returning to the neighborhood. One commented in June 2006 that,

The people start coming back to my place, my street. You know, my street is full. And my subdivision is too, quite a few people are back there, the majority of people are back in our subdivision so I would say that it's a plus, and we would see the trash picked up once a week, and the mail, we've been receiving that.

Conclusion

Our research contributes to geographical research on disasters and vulnerable populations by harnessing the disaster migration, social network, and place attachment literature. More specifically, this research fills a substantial gap in the disaster migration literature by examining the spatial morphology of evacuation routes, volumes, and frequencies, as well as the return rates and experiences. In addition, past research on the impacts of hurricane-induced and other natural disasters has focused primarily on the differential experiences of whites and African Americans, whereas this research compares African Americans and Vietnamese Americans, two minority groups residing in a specific, small, and spatially defined study area. Based on a post-Katrina comparative analysis of African American and Vietnamese American minority groups in one New Orleans suburban community, this research reveals substantial differences in the evacuation experiences and associated stresses of the return decision-making process.

Evacuation rates before Katrina made landfall were relatively high among both groups, although slightly lower for African Americans. These findings are consistent with Katrina research by Pastor and colleagues (2006). In addition, evacuation rates for females were slightly lower than for males, but African American females fared worst when compared to co-ethnic males and Vietnamese Americans in that a greater proportion evacuated by bus or other modes of transportation rather than by car.

The most significant contribution of this research is the focus on the spatial morphologies of evacuation, or evacuation geographies of Katrina, because no existing research calls attention to this aspect of the disaster evacuation experience. Our results reveal that African Americans experienced more moves than Vietnamese Americans, and that African American females experienced the highest frequency of moves. In addition, African American females were least likely to be evacuated with their families, less likely to have families or relatives in their first and second destinations, and remained in their destinations for a longer period. These results could reflect prevalent stressful life experiences perhaps due to racial bias and discrimination prior to and during Katrina, as well as a lack of access to resources. Although African American men felt greater degrees of discrimination during Katrina than pre-Katrina, they fared better in most measures compared to their female counterparts. In contrast, Vietnamese Americans possessed far more positive Katrina experiences in terms of less frequent evacuation moves because of a more defined regional social network of families and church to assist in the evacuation process.

Similarly, the role of social networks was instrumental in the very different return experiences of African Americans and Vietnamese Americans. When compared to evacuation experiences, however, explaining the differences in return experiences is far more complex. Much like McCarthy et al. (2006) determined for the Katrina diaspora in general, social networks of family and friends were influential for the volume and timing of the return decision-making process. The early return of Vietnamese Americans was anchored by the important role of Mary Queen of Vietnam Church in fostering this process and the deep family-based social networks that evolved among the refugee population that arrived more than thirty years ago. This contributed to alleviating the psychological stress among Vietnamese Americans in our study area, as well as to the rebuilding process we have described elsewhere (Chen et al. 2007; Leong et al. 2007; Airriess et al. 2008).

The synergistic role of social networks and place attachment influenced the return process as well. Overall, our empirical findings from both survey and focus group discussions confirm what has been found in existing research on place attachment and disaster: Both groups indicate their strong attachments to New Orleans, but mostly to their neighborhood, and such place attachment contributed to their decision to return. For elderly Vietnamese Americans, this neighborhood is the only place they have known since arriving more than thirty years ago. The second and now third generations have grown up together and families possess close relationships with one another. The parish church has long played a pivotal role in community life and this has only increased since Katrina. Such a sense of belonging and a strong social network among Vietnamese Americans, in particular, helped in mitigating the post-Katrina adjustment and return process. African American homeowners similarly possessed strong place attachment, but their attachments were tempered by less positive perceptions of the future of the community.

Our research examined both evacuation and return processes. However, the decision-making process regarding return is far more complicated. Whereas evacuation under conditions of an emergency situation inherently reduces decision-making choices, the return decision-making process entails a more complex set of choices regarding whether to return and the timing of that return. As such, there is a need to further tease out the complex reasoning associated with the return decision-making process. Elliot and Pais (2006) found that race aside, higher income households are able to delay the return process because of greater financial security. Our own fieldwork observations and informal interviews also suggest that higher household incomes among certain African American homeowners might have allowed them to remain longer in the Katrina diaspora.

Our study determined that social networks were important in the evacuation process in terms of evacuation destinations and types of assistance received, as well the decision to return or not, and when to return to the original neighborhood. The observed variations between the two minority groups in such processes largely resulted from different community social networks at various scales that had existed during the pre-Katrina years and the divergence between the two study groups since Katrina. Our research further demonstrates a need to develop more appropriate measures for disaster preparation, relief, and rebuilding, especially with reference to vulnerable population groups such as female-headed and linguistically isolated households, renters, and other economically disadvantaged groups. Only by taking serious steps toward such goals can we confront racial discrimination and economic injustice for the betterment of communities and society as a whole and, in the process, better prepare our communities for future natural or technological disasters.

About the Author

Her research interests focus on immigration and integration in the Pacific Rim and financial institutions and immigration.

His research interests include development, ethnicity, cultural landscapes, and East and Southeast Asia.

Her research focuses on mental and behavioral health issues among vulnerable populations, in particular ethnic minority and immigrant youth and their families.

Her research interests focus on the intersection of gender, race, and class.

Her research interests focus on the intersection of race, gender, and health.

Acknowledgments

*Two U.S. National Science Foundation grants (0555135 and 0555086) enabled the authors to conduct this project. The authors are indebted to all participants for their time and insights, especially to Cyndi and Thu Nguyen, Father Vien The Nguyen, and VIET volunteers in New Orleans and Houston; Maanulwa Mukule and the Multicultural Alliance in Houston; as well as all of the research assistants at Arizona State University for their invaluable assistance.

Notes

^a n = 14

^b n = 26

^c n = 53

^d n = 46.

¹. Our survey was conducted among residents in pre-Katrina ZIP code 70129, which includes the two-census-tract core study area. The rest of this ZIP code contains largely unclaimed and uninhabited swamps and marsh. For detailed sampling design and procedures, please see Chen, Keith, Leong et al. (2007) and Chen, Keith, Airriess et al. (2007).