Lasting Damages: Measuring New Orleans’ Property Value Recovery Rate

  1. Abstract: In 2005, Hurricane Katrina hit the gulf coast and became the costliest hurricane in U.S history (Knabb and Rhome, 2011). With broken levees and up to 19 feet of flooding in New Orleans, much of the city was affected. This property damage was said to affect some groups disproportionately with huge migrations out of the city. Using GIS, I studied the effects of this damage on property values five years after the hurricane and found that housing values had recovered and increased on average 1-2 percent and up to 10 percent from pre-storm levels, corroborating studies on the long lasting effects of out-migration after Katrina. 
  2. Background:

Hurricane Katrina, was the most destructive hurricane in U.S. history, with $108 billion in damage, up to 30 foot storm flooding, and 1,833 associated fatalities (FEMA 2015). The hurricane, which made landfall August 29th, 2005, was infamously poorly handled by a conglomerate of government agencies at the federal, state, and local levels (Stone, 2006; Shoup, 2005). 80% of New Orleans was flooded after the storm broke levees, and in the year following the hurricane, New Orleans’ population decreased 29% between 2000 and 2010 (Nodjimbadem, 2015). In a study by the U.S. Department of Housing and Urban Development’s Office of Policy Development and Research, it was reported that over 400,000 houses, out of 1 million, experienced minor to severe flood damage in Orleans parish alone (U.S. Department, 2006).

Many researchers have attempted to study the recovery process of this disaster in order to understand the potential direct and long-term inequities that the storm revealed and exacerbated. Studies have found strong evidence of permanent out-migration (Paxson and Rouse, 2008); and whiter, richer city demographics post-Katrina (Frey and Singer, 2006). Kates (2007) and Green et al. (2011) found that Hurricane Katrina exacerbated current population, economic, and infrastructure downturns, revealing and worsening existing inequities.

The immediate aftermath of the storm affected vulnerable populations differentially due to an inability to evacuate and weak infrastructure. Studies found that damaged neighborhoods were more likely black, and poor (Green et al., 2011). More than one-million of the 5.8 million hit hardest by Katrina were already impoverished, and 54% did not have a car (Shapiro and Sherman, 2005).  Paxson and Rouse, in their study The Impact of Hurricanes on Residents and Local Labor Markets: Returning to New Orleans after Hurricane Katrina, found that flooded property was the strongest indicator for probability of permanent emigration; African Americans were 45 percent more likely to experience flooding, compared to 25 percent of other residents. Even when controlling for this flooding, black residents were significantly less likely than other residents (49 percent) to return to New Orleans after the storm. The strength of these flooding results corroborates the use of flood data to measure damage in my study. The impacts of this out-migration is studied by Frey and Singer (2006) in their research, Katrina and Rita on Gulf Coast Populations: Census Findings. Using census data and special census bureau reports, the researchers found that post Katrina populations in New Orleans were in fact “more white, less poor, and more transitory than the pre-hurricane population,” (Frey and Singer, 2006).

In looking at the effects of Hurricane Katrina, these studies have provided evidence for the tangible social and economic consequences of uncoordinated disaster preparation and response. These studies are often framed around population numbers and socio-economic status before and after the disaster. In order to build upon this previous research, I will be examining property values as a proxy for the city’s recovery rates and dynamics. Property values extend these frameworks by addressing the question of infrastructure damage and reconstruction that structures the city’s availability and affordability for resident resettlement.

  1. Question: What is the spatial distribution and rate of Hurricane Katrina’s recovery?
  2. Methodology

In order to gain a big picture of the hurricanes affecting the Gulf Coast in 2005, I used the “all storms lines” data set on NOAA’s International Best Tracks Archive for Climate Stewardship. I then uploaded a basemap, selected the 2005 storms in the attribute table, and exported the layer. I symbolized the data according to wind speed and compared that to the Saffir-Simpson scale.

In evaluating recovery, 2000 and 2010 housing values from U.S. Census Factfinder are used as a proxy for the city’s recovery. The 2000 and 2010 census data sets are used, as opposed to years closer to the 2005 hurricane, because these years have comprehensive census surveys, as opposed to American Community Survey, with the most reliable and relevant data for this study.

These housing values are expressed over the 2010 TIGER census tracts. Census tract boundaries were used, as opposed to block groups or counties, in order to gain a relevant scope to view New Orleans’s distribution of recovery throughout the city. 2010 census tracts were joined with 2000 and 2010 median housing value census data from factfinder. Census tracts from 2010 were used for both housing value data sets because 2000 census tracts did not have a comparable geographical identifier to join with the housing data. After joining the data, I had to change the median value field to a numerical value to categorize it by adding a new “double” field and using field calculator with the new field equal to the original field name. Both panels of the figure use the same color scaling for house value. After this I categorized both the 2000 and 2010 data with the same bins. I used the 2010 bins given for both maps so that the changes in house values between the years would be visually comparable, but I added a sixth bin to include the lowest 2000 values so that the whole range was included. I also calculated the percent change in housing values from 2000 to 2010 according to equation 1, where C is percent change, V1 is median house value  for 2000, and V2 is median house value for 2010.

[equation 1] C=(V2-V1)/V2

In this equation I subtracted the median values of 2010 and 2000 and then divided by the 2010 value, and applied this equation to each census tract. This sheet was then joined with 2010 census tracts and the data was categorized to reflect the percent change.

Storm surge level was found using the NHC Tropical Cyclone report for Hurricane Katrina. The storm surge was recorded between 10-19 ft in New Orleans, and therefore an average of 15 feet storm surge was used. After uploading a digital elevation model of the area from the USGS LP DAAC, the elevation was converted from meters to feet using the times spatial analyst tool (1 meter=3.28 feet). After this conversion, the storm surge was found using the less than equal spatial analyst tool. The percent change data was then uploaded in an attempt to look at the two variables in relation. The storm surge map does not accurately depict inundation levels because of a faulty digital elevation model and unaccounted levee flooding.

Given your background discussion about poverty and equity, I feel like you need to provide some justification for the focus on housing values and changes in housing values.  You also need to justify the selection of 10-year time scale and tract-scale data. I’m excited to see the inundation map work out!

 

  1. Results:

Hurricane Katrina was just one of more than ten storms that hit the gulf coast in 2005 (Figure 1), but it was the strongest and the most destructive storm to hit land with up to 150 knot winds (Knabb and Rhome, 2011).

Figure 1: Hurricanes in the year 2005 off the Gulf Coast. Hurricanes, categorized by wind speed. Tropical depression: <38 mph, tropical storm:39-73 mph, Category 1:74-95 mph, Cat 2: 96-110mph, Cat 3: 111-129 mph, Cat 4: 130-156mph, Cat 5: >157mph. Rita and Katrina were the strongest storms to hit the Gulf Coast in 2005, though Katrina maintained Category 3 wind speeds for a longer period of time on land, causing more destruction.

Figure. 2: Hurricane Katrina storm track (following the eye of the storm) over Louisiana, Orleans Parish displayed in yellow. As the storm hit the Gulf Coast it decreased intensity from a category 5 out in the water to category 4 at the coast and category 3 around New Orleans. This decrease is due to a loss of energy as it moves away from the evaporation on the ocean’s surface and increasing friction under the storm as it reaches land (consistent with Iman et al, 2005). This storm track only shows the eye of the storm, which is why it seems to bypass New Orleans to the east. The storm’s effects expanded much wider than this eye track. The storm’s wind swath extended at least 75 mi out (Knabb, 2011). In addition, Category 4 and 5 winds in the Gulf created up to 55 foot waves, and produced 12 inches of rain (Knabb).

Figure 3a: Median housing values (in $) within census tracts, for 2000 and 2010. Zoomed in on downtown New Orleans.

Fig. 3b: Housing value for 2000 and 2010, broader view of surrounding Louisiana land. Same legend as Fig. 3a.

In looking at the effect of this storm, housing values within the census tracts, shown in Fig. 3a and 3b, show definitive changes between 2000 and 2010.  Five years after the storm, property values have increased over time. There is a significant increase in houses over $200,000 in 2010. Zooming in (Fig 3a), this trend is proven to be true downtown as it is for surrounding Louisiana lands (Fig 3b), especially concentrated in the coast and outside the city to the southeast. Comparing 2000 and 2010 values, it seems that areas with higher property values continued along that trend, while areas with relatively lower values (southwest corner and mid-east area in Figure 3a) remained that way. This is consistent with Kates’ (2007) study.  These changes are made clearer in Figure 4 below.

      

Fig. 4a and 4b: Housing Value Percent Change from 2000 to 2010

Looking at the percent changes of these values (Fig. 5), visually increasing property values in Figure 3a and 3b are corroborated. The tracts with decreasing value experienced (yellow) were concentrated in the downtown areas closer to Lake Pontchartrain, but not immediately on the lake property, and also to the east of the canal that ran in between the French Quarter and the Lower 9th ward. It makes sense that these areas were hit the hard by storm inundation. Lake-front property is more expensive, often with more affluent owners, so it makes sense that these areas were able to rebuild and increase value after the storm, while some dispersed tracts just inland of that were not able to reconstruct fully to their pre-storm levels. The lower 9th ward, home to a 98% black population, was strongly affected by a failing levee on the Mississippi River just directly to its west. To this day, population levels still have not recovered to their pre-storm levels, with only 37% population recovery in 2015 (Green et al., 2011; Allen, 2015).

Fig. 5: Flooded land with housing change

An attempt to analyze flooded areas in relation to these changes in housing value was not successful as New Orleans experienced much more inland flooding than the map suggests. The problem may have been caused by insufficient data (missing information about faulty levees), and a potentially inaccurate digital elevation model. However, this inaccurate storm surge map still points to interesting spatial information for this area. Specifically, the significant effect of the city’s failed levees on storm surge is evident with the apparent lack of flooding when this is unaccounted for.

  1. Discussion:

Hurricane Katrina created catastrophic hurricane property damage and mass migration out of the city (Paxson and Rouse 2008; Green et al., 2011). With high wind speeds, far-reaching inundation, fatalities, and damaged infrastructure, the response and reconstruction after Katrina was a challenging undertaking. Five years after the disaster, property values, displayed in Figures 3 and 4,  show that housing values have actually increased over time, suggesting some sort of economic recovery and rebuild since the storm. This overall increasing value trend supports existing research on permanent emigration of poor and black residents out of New Orleans, since the observed housing reconstruction seems to leave these groups out. Figure 5 does show some areas of decreased housing value, mostly concentrated just inland of lake-front properties and west of the failed Mississippi River levee; Figure 3a shows the maintenance of areas with relatively lower values compared to relatively higher values. This trend corroborates studies (Green et al., 2011; Kates et al., 2007) that prove disasters exacerbating existing inequities.

It would have been more illustrative to use census data from years closer to either side of the 2005 storm, but housing at a census tract level was not available for the American Community Survey. The census tracts for 2000 would also improve the 2000 housing value data, possibly filling in more missing tracts. Unfortunately the census tract data set for this year did not include a shared field (i.e. geoid) with housing values from the same year. The storm surge must be made more accurate to draw any definitive conclusions from it.

Despite difficulty directly linking hurricane impact and housing values after the storm, the results were instructive in that they showed an overall increase in housing values five years after the storm. Reconstruction of the city seems to be increasing property values, potentially due to the ability and willingness of more affluent residents to rebuild, and the permanent out-migration of poorer residents who would occupy those lower-value homes. Accurate storm surge data would be very helpful to support a connection between this result and direct storm damage.

To improve this study I would like to look at the changes of income and or race demographics before and after the storm, because many studies suggest that the storm’s effect on these factors have still not recovered to their pre-Katrina levels. I would like to see if there was a significant spatial component to the concentrations of theses shifts. This spatial data would strengthen results of the housing data to support researchers claims that recovery of Hurricane Katrina has left out the poorest, black members of New Orleans’ community.

  1. Bibliography

Allen, Greg. August 3, 2015. “Ghosts of Katrina Still Haunt New Orleans’ Shattered Lower Ninth Ward”. National Public Radio. http://www.npr.org/2015/08/03/427844717/ghosts-of-katrina-still-haunt-new-orleans-shattered-lower-ninth-ward

Federal Emergency Management Agency (FEMA). 2015. “Hurricane Katrina Overview”.  https://www.fema.gov/hurricane-katrina-overview

Frey, William and Audrey Singer. 2006. “Katrina and Rita on Gulf Coast Populations: Census Findings.” The Brookings Institution.

Green, R., Kouassi, D., & Mambo, M. (2013). Housing, Race, and Recovery from Hurricane Katrina. The Review of Black Political Economy, 40(2), 145-163.

Iman, R. L., Johnson, M. E. and Watson, C. C. (2005), Sensitivity Analysis for Computer Model Projections of Hurricane Losses. Risk Analysis, 25: 1277–1297. doi:10.1111/j.1539-6924.2005.00673.x

Kates, R. W., C. E. Colten, S. Laska, S. P. Leatherman, and William C. Clark. 2007. “Reconstruction of New Orleans after Hurricane Katrina: A Research Perspective.” Cityscape 9, no. 3 : 5-22.

Knabb, Richard, Jamie R. Rhome, and Daniel P. Brown. 2011 “Tropical Cyclone Report: Hurricane Katrina.” National Hurricane Center.

Nodjimbadem, Katie. 2015. “These maps show the severe impact of Hurricane Katrina on New Orleans.” Smithsonian Magazine. http://www.smithsonianmag.com/history/these-maps-show-severe-impact-hurricane-katrina-new-orleans-180956364/

Paxson, Christina, and Cecilia Elena Rouse. 2008. “Returning to New Orleans after Hurricane Katrina.” American Economic Review 98, no. 2: 38-42. EconLit with Full Text, EBSCOhost

Shapiro, Isaac and Arloc Sherman. 2005. “Essential Facts about the Victims of Hurricane Katrina.” Center on Budget and Policy. http://www.cbpp.org/research/essential-facts-about-the-victims-of-hurricane-katrina

Shoup, Anna. Sep 9, 2005. “FEMA Faces Intense Scrutiny.” PBS Newshour. http://www.pbs.org/newshour/updates/government_programs-july-dec05-fema_09-09/

Stone, John D. 2006. “Issues MISread, Risks MIScalculated, and Crises MISmanaged: Public Relations Lessons Learned From The Catastrophes of Katrina.” Public Relations Quarterly 51, no. 4: 36-39. Business Source Premier, EBSCOhost

U.S. Department of Housing and Urban Development’s Office of Policy Development and Research. 2006. “ Current Housing Unit Damage Estimates.” Federal Emergency Management Agency. https://gnocdc.s3.amazonaws.com/reports/Katrina_Rita_Wilma_Damage_2_12_06___revised.pdf.

  1. Appendix
  1. Data Sources:

Hurricane track data was find using NOAA’s International Best Track Archive for Climate Stewardship (IBTRaCS), all storms lines dataset, found here. The median home value data sets for 2000 and 2010 were found on the U.S. Census Bureau’s American FactFinder databases. Census tracts were found on TIGER, though 2000 census tracts were not usable to join with housing. The neighborhood boundaries, used in Figure 4, were collected from Zillow’s Neighborhood Boundaries data set: (https://www.zillow.com/howto/api/neighborhood-boundaries.htm)

The digital elevation model is from USGS global data explorer site (https://gdex.cr.usgs.gov/gdex/). To find DEM data, make sure ASTER Global DEM is checked for background image and ASTER Global DEM is checked for data coverage. Then you are able to define a rectangular area in the toolbar and download the TIFF data.

http://web.archive.org/web/20120313012408/http://www.fema.gov/hazard/flood/recoverydata/katrina/katrina_la_gis.shtm

FEMA Hurricane Recovery Flood Maps

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s