ABSTRACT
Flooding in the United States impacts over 40 million people, disproportionately affecting low-income and non-white communities, while costing at least $8 billion annually. Large scale response-based plans have become popular adaptation tools; however, there are limited analyses of their effectiveness and equity. The State of Louisiana’s 2017 Comprehensive Master Plan for a Sustainable Coast outlines billions of dollars worth of investments into coastal restoration and adaptation projects with the aim to protect the unique communities and cultures of coastal Louisiana. We reviewed the Plan’s storm surge flood-depth modeling to determine the expected beneficiaries of implemented projects and the degree to which the Plan achieves its goals of protecting unique communities and cultures. We constructed two descriptive statistical tests of Master Plan data. We find the Plan is effective at reducing flooding across the region, relative to a future without it. However, racialized populations within areas with high levels of social vulnerability are disproportionately negatively affected or not affected at all. We conclude that the Master Plan is unevenly effective and equitable for the diverse communities of coastal Louisiana.
Key planning implications
Louisiana’s 2017 coastal Master Plan fails to fully address existing flood hazards.
Where the Master Plan does address flood hazards, it does so unevenly, across existing lines of vulnerability and race. The Master Plan reduces risk for some people but not all, creating winners and losers from adaptation projects.
Our findings highlight a persistent issue where the lack of analyses pre- and post-implementation of adaptation plans reproduces maladaptive patterns.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Competing interest
The authors have no relevant financial or non-financial interest to disclose.
Author contributions
Michael Molloy drafted the manuscript, designed the research, gathered data, conducted methods, and interpreted results.
Dr. Eric Nost designed the research, conducted methods, interpreted results, revised the manuscript, drafted the manuscript.
Megan Bledsoe drafted the manuscript and provided editorial services through each iteration of the manuscript.
Notes
1 Our areal weighting approach produces some inaccuracies since populations are not always evenly distributed across Census blocks. However, we believe this approach is adequate since blocks are the Census’s most fine-grain measure of population. Separately, we tested our approach using the US Environmental Protection Agency’s relatively fine-grain 30m x 30m grid of estimated population (see Flores et al., Citation2022; Wing et al., Citation2022) and did not find any significant differences in our results. Future iterations of our analysis could attempt to estimate the distribution of people even more precisely across the coast using ancillary data such as satellite-observed nighttime lights or building footprint polygons.
2 We breakdown the results from each test in terms of social vulnerability category as well as racial demographics. We do this because both vulnerability and race are important but separate considerations in any adaptation policy and planning analysis. In our case, we ask whether the Master Plan disproportionately affects communities by vulnerability status and whether there are racialized disparities in its predicted outcomes. We do not use measures of race as indicators of social vulnerability, except insofar as CPRA itself does.
3 In joining the data from a coarser resolution (Census block group) down to a finer resolution (Census block), we may be, in some cases, making an ecological fallacy. The centroid of a large, populated flood area would fall within just one of several Census block groups it spans, and it would not necessarily be appropriate to determine the vulnerability of that entire area based on just one block group. However, a visual analysis indicates that there is generally a good approximation between populated flood-prone areas and block group boundaries (e). An alternative would be to summarize flood depths at the level of the block group, in terms of median, mean, or maximum difference (Test 1) and change (Test 2). However, this approach would wash out potentially important variabilities in flooding within a block group. We determined that risking an ecological fallacy – imputing a social vulnerability score determined for a larger area, to a smaller area - was more conservative in terms of not under-estimating risks.
4 Our entire analysis can be viewed and reproduced here: https://colab.research.google.com/drive/1IHHu_1v_WTaidVvHwWnc2QvXBiPDs2ao?usp=sharing