In this project, we use GIS to evaluate the community’s post-disaster social fabric condition associated with the Federal Emergency Management Agency’s (FEMA) home buyout program. We found that while the home buyout program is able to relocate people after the flood hits the community and decrease the social vulnerability from an individual’s perspective, it could also bring unprecedented harm or negative impact to a community from a community’s social fabric perspective. We insist to think more research is neccessary regarding evaluating the rationality of policies regarding increasing community resilience. To go to the PDF version of this poster, here it is!
7 thoughts on “Assessing Community’s Post-Disaster Social Fabric associated with Home Buyout Program”
Comments are closed.
Hi there! Thanks for visiting my poster. This project is inspired by one of the biggest flood hazards in Tennessee history-Nashville flood in 2010. While many research has been conducted on understanding the advantages and disadvantages of community resilience policies, such as the home buyout program or construction elevation plan, most of them focused on the perspective of how these policies would impact an individual’s vulnerability rather than on a community’s social fabric point of interest. Our preliminary result shows that FEMA’s home buyout program could unprecedentedly impact a community’s social fabric condition and lead to a community’s social fabric breakdown despite the fact that it can relocate residents affected by flood to new locations.
I would like to hear your comments and feedback.
Bowen, this is super interesting! My family moved to a city just south of Nashville just before the flood, so I was here to experience it. Regarding your analysis, I was wondering why you chose 2018 as your year for modeling social fabric after the flood – does it have something to do with the timeline of the buyout program? Do you think your analysis would show a worse scenario of social fabric in a year prior to 2018 yet still post-disaster?
Hi Patrick! Thanks for your comment. Essentially I was just trying to collect data after the flood and the data should be as latest as possible to show the impact of the home buyout policy on the community’s social fabric since the impact of the policy needs time to reflect on the data. However, I couldn’t find complete census data for 2019 or 2020 so I just make some compromise to use 2018 census data. I think my analysis would show a better scenario of the social fabric in a year prior to 2018, and the scenario should be turning better when the data comes closer to pre-event (2010). Thanks!
Bowen, this is a great topic, and I love the local connection. I had no idea the social fabric score exists, so thanks for introducing me to that. I am curious about what drove the difference in home buyout rates in the two census blocks you focus on. Did the control block see as much flood damage as the adjacent block? If so, do you know why the control saw fewer buyouts? If not, could the difference in flood damage itself contribute to the difference in the social fabric scores?
Hi Colton! Thanks for your comment. The control census block didn’t see much flood damage as the adjacent block so the home buyout rates are very low. And this is also the reason why I choose this census tract as control census tract. We can understand that the spatial distribution of flood damage itself contributes to home buyout distribution, and further controls the spatial distribution of social fabric score/condition.
Bowen, excellent work all around–from generating your social fabric index and PCA results as measures, and you have a very well designed natural experiment here, comparing adjacent tracts. It looks like a strong effect and worthy of pursuing for publication I believe; I encourage you to pursue it further. I wonder how complex it would be to apply this methodology to different cities/contexts? A publication that laid out the steps for application elsewhere would be a nice contribution, enabling comparability and reproducibility of results in different contexts. Great jobn.
Thanks Steve! In order to apply this methodology to different cities/contexts, the biggest problem will be the data availability associated with that context. For example, the social fabric score could include lots of information regarding the points of interest in the historical time such as how many churches or schools in a census tract/block in 2010 and back in 2010. This information is not easily available across the country, even if in GIS database or opening street maps. One possible solution would be the spatial bayesian multilevel modeling with R-INLA package, using a generalized linear model with a Poisson distribution as the link function to predict/assess the points of interest in our context of interest. Future work could be incorporating some of the community survey data into our social fabric score to incorporate people’s objective feelings into our consideration. The missing data regarding the community survey could also be predicted by the spatial multilevel modeling with the bayesian inference technique. Thanks!