Katrina Study (PKSRR)Workshop on Longitudinal Study of

Natural DisastersHarvard Population Center

June 11, 2011

Page 2

Adversity and Resilience Study

Christina Paxson, Economics Princeton Mary C. Waters, Sociology Harvard Jean Rhodes, Clinical Psychology U Mass Boston Elizabeth Fussell, Sociology Washington State Cecelia Rouse, Economics Princeton Sarah Lowe, Psychology, U Mass Boston →Columbia Christian Chan, Psychology, U Mass Boston→ University of Hong Kong Corina Graif, Sociology, Harvard→ University of Michigan Nicole Deterding, Sociology, Harvard Margaret Willis, Sociology, Boston College

- Funded by NICHD, MacArthur Foundation, National Science Foundation

Hurricane Katrina 2005

Timeline- Hurricane Katrina came into the Gulf on August 25, and increased to

a Category 5 (winds peaking at 175 mph) by August 28.- Mandatory evacuation of New Orleans ordered on August 28. - Estimates are that ~1M (of ~1.2M) residents evacuated.- Landfall in Louisiana: August 29. Category 3 with winds of ~125 mph.- Extensive wind damage.- 53 levee breaches produced extensive flooding: - 80% of the city of New Orleans was flooded as of September 2, with

water levels reaching 20 feet.- It took weeks to pump the city out. - Both a “natural” and “man made” disaster.

Hurricane Katrina

Overall effects:- Total costs estimated to be $81.2 billion.- $30 billion in Federal aid.- 1836 deaths, majority in Louisiana.- 90,000 square miles declared a disaster area (Equal to the entire land

mass of the UK)- Displaced 650,000 people- Destroyed 217,000 homes- 60% of housing stock in New Orleans city was destroyed- 30% of housing stock in New Orleans MSA was destroyed

New Orleans City lost 29.1% of its population between 2000 and 2010. (Detroit lost 22.2%)

Opening Doors Study

Launched in 2003 by MDRC, MacArthur Network on Transition to Adulthood Designed to test whether tuition supplements, enhanced student services or

curricular enhancements could affect retention of students in community colleges.

Experiments conducted in 6 community colleges across 4 states. Chris Paxson added questions on physical and mental health. Louisiana Site—2 Community Colleges

- Delgado Community College- Louisiana Technical College West Jefferson

Louisiana Opening Doors Study

Funding Came from Louisiana Department of Social Services and Louisiana Workforce Commission. TANF funds paid for the program but individuals did not need to qualify for TANF to be part of the study.

To qualify for the study students had to be- Between the ages of 18-34- Parent of at least one dependent child under the age of 19- Family income < 200% of poverty level- GED or High School Degree- No degree or occupational certificate from an accredited college or

university program

The Opening Doors Sample N=1019, at baseline

92% female 85% black 19% married Average age 26 98% ever worked 71% receiving government benefits. 52% currently employed 43% first in family to attend college 69% had access to a working car Average age of children 3 years A disproportionate number come from the 9th Ward.

12 Month SurveySample A12/04-8/05

N=492

Baseline Survey 11/03-2/05

N=1019

Post Katrina SurveySample A5/06-2/07

N=402Response Rate 82%

Post Katrina 12 MonthSample B3/06-2/07

N=309Response Rate 58%

Hurricane Katrina 8/25/05

Second Follow UpSpring 09-10

Samples A and B1019 eligible

N=720Response Rate 70.6%Genetic Study N=270

Qualitative Interviews

N=57

Qualitative Interviews

N=63

Qualitative Interviews N=120

First Wave Conducted after the 2006-2007 survey and linked to survey responses (57 interviews)

Second Wave Conducted after the 2009-2010 survey and linked to survey and previous qualitative responses. (63 interviews)

Equal number of people who were back in New Orleans, and who had relocated to Texas.

Covered Hurricane Experiences, life history, politics, intergroup relations, experiences of young adulthood, and questions about how their children are doing.

Transcribed and coded using Atlas Ti. Interviews were linked to the longitudinal survey data to

contextualize the interviews and to give a rich understanding of the trajectories of individuals.

Advantages of Our Sample

Most studies of disasters do not have data on people before the disaster. We had two waves of data on Sample A and one wave on Sample B before the hurricane. Our data included - physical and mental health- economic resources - social support- social trust- future aspirations and expectations- measures such as optimism, self esteem, confidence

Disasters have unequal impacts, generally exacerbating inequality and differentially affecting women, the poor, and racial minorities

The scope of Katrina makes it a very unusual and important disaster Disasters are seldom studied longitudinally. We really do not know

a lot about long term recovery.

Longitudinal Data on Resources and Outcomes

Studies to Date

College Re-enrollment Pet Loss Pre-disaster social support Child-related stressors Natural mentors Intimate relationships Interviewer race Decisions to evacuate Resilience Trajectories Children’s Functioning Combining Quantitative and Qualitative Methods

Religion Barriers to Community College Completion Geographic Mobility Relocation Decision Making Neighborhood Attainment Employment Trajectories Changes in BMI Post Traumatic Growth Conservation of Resources Theory Transition to Adulthood

The Overall Picture

Psychological Resilience—defined as a return to pre-disaster levels of psychological functioning is the normal reaction to a disaster, even though it is remarkable.

Post traumatic growth is also widespread. Defined as subjective psychological gains directly related to the trauma they endured.

There is a lot of research on psychological resilience, very little on how social and economic resistance is related to it.

We know that community ties and social networks as well as socioeconomic resources are required for social well being. How does this affect recovery?

Recovery in New Orleans overall has been market driven, with federal money distributed on an individual basis.

New Orleans is recovering—but it is a changed city. Post Katrina it is smaller, older, more educated, less poor, fewer renters, fewer households with kids. New Orleans has its lowest poverty rate since 1979.

People who did not return are more likely to be poor, African American households with children.

Dilemma that they face: better individual opportunities outside New Orleans vs. the sense of community they had.

A false dilemma?

Katrina Traumas

In the week after Hurricane Katrina hit was there a time when you:

Katrina Traumas Percent

Did not have enough fresh water to drink 26

Did not have enough food to eat 35

Felt your life was in danger 32

Didn’t have medicine you needed 32

Needed medical care and couldn’t get it 30

With a family member who needed medical care and could not get it. 33

Didn’t know if child/children were safe 23

Didn’t know if other family members were safe 77

Were any of your relatives or close friends killed because of Hurricane Katrina or Rita? 31

Mean # Katrina Traumas 3.14

Trauma Exposure• 80.8% experienced home damage • 32.1% experienced the death of a

friend or relative (Paxson, et al. 2012.) • Rise in domestic violence and stressed

relationships with partners, even among people who had not experienced this before (Lowe, Rhodes, & Scoglio, in press).

Psychological and Physical Distress WorsenedName Pre K Measure Post K Measure

K6 Scale (range 0-24)(Psychological Distress)

5.63(4.23)

6.65***(5.38)

K6 Scale Over 7(Probable Mild to Serious Mental Illness)

0.233(0.42)

0.371***(0.557)

K6 Scale Over 12(Probable Serious Mental Illness)

0.069(0,23)

.138*(0.363)

Perceived Stress Scale (0=low stress to 16 high stress)

4.31(3.31)

5.37***(3.87)

Fair or Poor Self Rated Health Status0.126

(0.332)

0.188***(0.444)

Number of diagnosed health conditions1.34

(1.28)

1.73***(1.40)

Body mass index (kg/m2)28.30(6.48)

29.13***(3.01)

K6(z-score)

MMI/SMI SMI IES-R

(z-score) PTSD PSS(z-score) FPH HC BMI

No additional controls

Housing or property loss “enormous”

0.283***(0.101)

0.098**(0.049)

0.078**(0.036)

0.346***(0.096)

0.175***

(0.049)

0.139(0.103)

0.070*(0.040)

0.196***

(0.141)

0.742(0.714)

# Katrina traumas 0.083***(0.020)

0.041***(0.010)

0.017**(0.007)

0.125***(0.0190

0.053***

(0.010)

0.064***(0.021)

0.023***

(0.008)

0.137(0.028)

0.300**(0.143)

Associations between Trauma and Outcomes Without Baseline Controls

There are large and significant associations between housing or property loss and nearly every outcome. Individuals who reported large housing or property loss are 9.8% more likely to have K6>7—probable mild to serious mental illness), 7.8% more likely to have K6>12 serious mental illness, and 17.5% more likely to have likely PTSD. They are also more likely to report fair or poor health, and greater numbers of health conditions.

Traumatic experiences associated with the hurricane are also significantly associated with nearly every health measure.For example, a person with the average number of traumas (3.14) is 5.3% more likely to have SMI than someone with no traumas and 18% more likely to have IES-R score>1.5 (likely PTSD)

How does the inclusion of pre-hurricane measures affect inferences?

The inclusion of socio-demographic controls has remarkably little influence on the coefficients

The associations between loss and trauma and health outcomes can be partially but not fully explained by pre-disaster health status.

- The inclusion of controls for pre K health reduces the coefficients for losses and trauma on MMI/SMI by approximately 20%.

- Coefficients for perceived stress and physical health generally decline as well.

- But the results for the two measures of PTSD symptoms change little with the inclusion of pre K health.

(Controls include: age, race, #children, income, receive benefits, car, social support scale and number of months between hurricane and survey to capture systematic changes in health that occur over time.)

- - Controls include: age, race, #children, income, receive benefits, car, social support scale

and number of months between hurricane and survey to capture systematic changes in health that occur over time.

- Controls include: age, race, #children, income, receive benefits, car, social support scale and number of months between hurricane and survey to capture systematic changes in health that occur over time.

Conclusions on Health and Katrina9-19 months after the storm at First Follow Up

Prevalence of serious mental illness doubled from 6.9% to `13.8% There was an increase in self reported poor or fair health, and both

overweight and prevalence of having at least one diagnosed medical condition rose significantly.

Nearly half (47.9%) of respondents exhibited probable PTSD post Katrina. Higher levels of loss and trauma were generally associated with worse

health status, controlling for baseline socio-demographic and health measures.

Higher baseline resources predicted less hurricane associated trauma, but the consequences of trauma and loss were similar regardless of baseline resources.- For instance, if you had a working car and money for a hotel you were more

likely to evacuate and experience less trauma, but if you did not evacuate those higher resources did not prevent you from experiencing negative consequences.

Five Years After 2009-2010

Second follow up happened between March 2009 and August 2010- Lived in 23 states- 16% of them were in their pre-hurricane homes.

What was the trajectory of mental health over the two survey points?

Trajectories of Recovery 4 years after the storm: PTSD

Impact of Event Scale Revised IES-R>1.5

Resistant (44%)- Did not show PTSD at either first or second follow-up

Recovered (23%)- Had high PTSD at first but not in second follow-up

Delayed (10%)- No symptoms at first follow-up, but present in - second follow-up

Chronic (23%)- High PTSD in both follow-ups

The IES-R is a 22-item self-report scale that assesses distress produced by a specific traumatic event. It was developed to cover all DSM IV criteria for post-traumatic stress disorder (PTSD) The IES-R asks respondents to rate how much they were bothered by symptoms of distress about a traumatic event over the previous 7 days. Responses are coded from 0 (“never”) to 4 (“extremely”.) Scores are averaged over items, producing a total score that ranges from 0 to 4. Earlier research indicates that individuals with IES-R scores exceeding 1.5 are at high risk of having PTSD.

67% are doing well

33% are suffering

Trajectories of Recovery 4 years after: Psychological Distress (PD)

K6 Scale >7

Resistant (53%)- Did not show PD at either first or second follow-up

Recovered (17%)- Had high PD at first but not in second follow-up

Delayed (12%)- No symptoms at first follow-up, but present in - second follow-up

Chronic (18%)- High PD in both follow-ups

. The K6 is a 6-item self-report scale that measures non-specific psychological distress and is used as a screener for anxiety and mood disorders. It is widely used in survey research and has good psychometric properties. The items in the scale ask about feelings such as “hopelessness” experienced during the past 30 days, with responses coded from 0 (“none of the time”) to 4 (“all of the time“.)

70% are doing well

30% are suffering

Three Measures of Trauma and Loss

Whether the respondent’s home experienced hurricane related damage.

Hurricane Trauma scale - 8 questions addressing experiences in the week following

the hurricane

Whether a close friend or family member died as a result of the hurricane

What predicts high PTSD in 2009-2010?

Trauma and loss associated with the hurricane.- The effects of home damage and each additional trauma on the odds of

having high PTSS increased from 2.48 to 4.76 and from 1.10 to 1.36, respectively, between the two surveys.

- The effect of having a family member or friend die due to the hurricane decreased from 2.13 to 1.55.

Mental health problems before the hurricane increased the odds of high PTSD by 1.92 and 1.81 respectively at the first and second follow ups.

Only two sociodemographic factors predicted PTSD: - older women were more likely to experience it than younger women- Those who earned more before the hurricane were more likely to

experience PTSD.

What predicts high Psychological Distress in 2009-2010?

Only traumas and death predict Psychological Distress, not home damage.

Those who had greater levels of social support at baseline were significantly less likely to show high psychological stress at either follow up.

Post traumatic growth5 subscales

Relating to Others- “I have a greater sense of closeness to others”

New Possibilities- “I developed new interests”

Personal Strength- “ I have a greater feeling of self reliance”

Spiritual Change- “I have a stronger religious faith”

Appreciation of life- “I have a greater appreciation for the value of my own life”

Post-traumatic growth

• PTG was found to be strongly positively associated with symptoms of PTSD • Only those participants with high levels

of PTSD at both time points maintained high levels of PTG over time (Lowe, Manove, & Rhodes, 2012).

Religion and PTG• Pre-disaster religious involvement and

faith were predictive of better post-disaster social resources which, in turn, were associated with lower levels of psychological distress (Chan, Perez, & Rhodes, 2010).

• Religious coping affected post-hurricane outcomes (Chan, Perez, & Rhodes, 2012).

Child Functioning• Concerns about child welfare affected

maternal mental health (Lowe, Chan, & Rhodes 2011).

• There were strong associations between child externalizing and internalizing symptoms and maternal psychological functioning (Lowe, Godoy, Carter, & Rhodes, 2012).

Pathways

Mobility• Only 19% of our sample members reported no

moves between baseline and all waves of the survey

• 58% of participants had changed tracts between baseline and 2006

• 63% changed tracts between 2006 and 2009. • Individuals who were home owners and whose

homes had not been flooded were more likely to return to NOLA (Paxson and Rouse, 2008).

Leaving New Orleans

SB How worried were you during that time when you were evacuating? R I didn’t panic. I was kind of a little shaken up because this was all

new to me. We were on a bus and tears just began to run out of my eyes because I was going to now an unfamiliar place that I have never been. I’ve been in New Orleans all my life. All my children, except for the little baby, were born and raised in New Orleans. My entire family, that is where we are from. New Orleans. So it just hurt me just to see the stage that the city was in. And to know that it would never be the same. Because while we were riding, I was just looking. Everything looks abandoned.

Figure 4: (A) Spatial distribution across the U.S. of New Orleans respondents one to two years after Katrina.

Distribution of individual applications for assistance from FEMA in 2007 at the Metro area level.

New Orleans has changed in fundamental ways

New Orleans has changed

July 2005 Population 455,188 July 2009 Population 354,850 (78%) City Schools are at 61% of their pre Katrina enrollment Lower Ninth Ward (22%) Metro Area is at 91% of its 2005 pre Katrina population Average wages increased 14% between 2004 and 2008 Latinos increased from 3% in 2000 to 4.7% in 2009 (probably undercounted) Blacks have decreased from 67% in 2005 to 61% in 2009 Violent crime and property crime have risen since Katrina

Arne Duncan, Secretary of Education

“Let me be really honest. I think the best thing that happened to the education system in New Orleans was Hurricane Katrina. That education system was a disaster and it took Hurricane Katrina to wake up the community to say that “we have to do better”. And the progress that they’ve made in four years since the hurricane is unbelievable.”- January 29, 2010

Neighborhood CharacteristicsTract Level 2000 Census Data

The respondents are on average living in significantly better neighborhoods in terms of poverty and disadvantage. They are also living in significantly less segregated neighborhoods.

The new neighborhoods have less people in poverty (22% vs. 26%), less people with public assistance income (11% vs. 14%), less female headed households with children (40% vs. 46%).

They also have more college graduates (18% vs. 15%) and more professionals (16% vs. 12%).

These neighborhoods are more white (39% post K vs. 30% pre K), less black (48% vs. 63%) and nearly double the number of Hispanics (9% vs. 4%).

30

39

62

48

2622

1410

4237

0

10

20

30

40

50

60

70

White Black Poverty PubAssist

UnempMales

Pre KPost K

Census Tract Characteristics Pre and Post KatrinaAll Respondents N=867

Flooding and Moving, in 2006-07

Of the non flooded respondents, those who move end up in slightly less poor neighborhoods.

Of the flooded respondents, they show considerable improvement in the quality of their neighborhoods.

Respondents who have moved outside of Louisiana are in much better neighborhoods.

12

39

82

44

34

20 1710

48

34

0

10

20

30

40

50

60

70

80

90

White Black Poverty PubAssist

UnempMales

StayersMovers

Census Tract Characteristics Stayers vs. MoversFor those who Experienced Flooding N=352

23

44

71

30 29

15 16

7

45

29

0

10

20

30

40

50

60

70

80

White Black Poverty PubAssist

UnempMales

Pre KPost outside LA

Census Tract Characteristics Pre and Post KatrinaFor those Moving outside Louisiana N=239

Do They Like the New Neighborhoods?Themes from the Qualitative Interviews

Many people report better job opportunities, less crime. Every person who moved out of New Orleans said the schools were better wherever they ended up.

The new diversity scares a lot of people. Change is experienced as exciting and filled with possibilities and as terrifying.

Many people report that they feel discriminated against in new locations because they are from New Orleans.

Most people miss their strong ties to family and neighbors and street life in New Orleans. For some that pulls them back even if they think they would be better off somewhere else.

P27: SB51795250. INTERVIEWER: What were the pros and cons? RESPONDENT: If I go back to New Orleans, I know that

economically it’s not good right now because everything is just not functioning properly right now. So that was going to be bad. If I stayed out here, things were looking up as far as job-wise, like opportunity-wise, but my family’s not here. It’s boring. I’m lonely. It doesn’t even feel the same at all. It just feels like I’m just here. So I can go somewhere here, but no one is here with me. But New Orleans is fun. I know people. I’m comfortable, but I’m not going to get anywhere. I’m like -- so that’s my little battle.

The dilemma of whether to return.

Trajectories of functioning

• Trajectories of l functioning: The multi-wave data has made it possible to investigate trajectories of psychological functioning, social support, and employment using Latent Class Growth Analysis (LCGA)

• We have also drawn on the multiple waves of data to demonstrate how post-disaster cross-sectional estimates of the impact of traumatic stress exposure produce somewhat inflated estimates of disaster and social support effects (Lowe, Greene, & Rhodes, in press).

Neighborhood change

Social Support

Other Findings

Most people report personal and spiritual growth from the experience. Those who blamed God or who thought that God was punishing them were the most psychologically distressed four years later.

For people with low social support at baseline, pet loss was the most significant predictor of psychological distress, and for many it was long lasting.

People who had high Psychological Distress at baseline significantly overestimated flood depths, relative to geocoded data.

Optimism was one of the best predictors of who did not evacuate.

Current studies• Exposure meta-analysis• PTG as a personality construct• Disaster and Health (BMI)• Legal issues, housing, etc. • Community College students