Since we’re currently in the middle of tornado season and on the heels of a prediction of an active 2018 Atlantic Hurricane Season, this would be an appropriate time for a discussion on natural disasters.
I’ve seen first-hand the destruction of natural disasters. I experienced Hurricanes Fran and Lloyd, plus the 2011 Super Outbreak of tornadoes while living in North Carolina and Alabama, respectively. It doesn’t take long after surveying the debris and downed trees or hearing the stories of survivors to realize the emotional and physical impacts of these events. But one thing that’s often talked about but harder to pin down is the economic impacts from nature’s fiercest displays of force.
That being said, let’s pose the question: “What results and trends can we expect when a natural disaster event occurs in a given region?” Seems simple enough, right? Well, the answer might be a little more complex than you’d expect.
A Recipe for Disaster
If you’re a music lover and have ever come up with a themed playlist, or just an economic researcher, you know you have to set some ground rules on criteria.
Mark Twain famously said: “The difference between the almost right word and the right word is really a large matter. ’tis the difference between the lightning bug and the lightning.” I think it’s also the difference between the nursery rhyme “Row, Row, Row Your Boat” and Hues Corporation’s “Rock the Boat” being played at a lake party.
“What’s wrong, I thought you said ‘boat’ songs!?” Don’t be that guy.
Anyway, let’s set ourselves up for success. The term natural disaster is defined by Merriam-Webster as, “a sudden and terrible event in nature (such as a hurricane, tornado, or flood) that usually results in serious damage and many deaths.” That’s a great start, but still a little too vague for this context. In another poor analogy, we need criteria that will help decide whether Talking Heads’ “Burning Down the House” or Scorpions’ “Rock You Like a Hurricane” belong on our ‘Natural Disaster’ playlist.
For the criteria we need, let’s refer to a working paper released by the National Bureau of Economic Research (NBER) entitled, “The Effect of Natural Disasters on Economic Activity in U.S. Counties: A Century of Data” by the authors Leah Platt Boustan, Matthew E. Kahn, Paul W. Rhode, and Maria Lucia Yanguas.
The authors of this research constructed a 90-year panel data set that included the universe of natural disasters in the United States from 1920 to 2010. Then, they divided disasters into two main groups: 1. A severe disaster resulting in 10 or more deaths, and 2. A “super-severe” disaster resulting in 100 or more deaths.
|The map above, courtesy of Alert Systems Group, overlays regional natural disaster risks based on data from The Red Cross and NOAA.
Additionally, all human-related disasters—such as mine collapses, transportation accidents, and arson—were excluded. So let’s skip on adding that Talking Heads track to the playlist.
From this framework, we can dive into the results of what we consider natural disasters and account for the variance in large-scale national events and the more regional ones. For the purposes of this blog, let’s focus on the more regional disasters.
“Watch out for that first step. It's a doozy!"
Did you catch the Groundhog Day reference? Besides being unimaginative, it echoes the trend of the results whenever natural disasters occur time and time again.
Continuing on with the NBER paper mentioned before, the study found that mild disasters had relatively marginal effects on the local economy. However, counties hit by severe disasters experienced increased rates of outmigration, lower home values, and higher poverty rates.
The research also touches on some of the socio-economic impacts by noting the correlation between outmigration and poverty rates. As outmigration of the more affluent population in areas hit by super-severe disasters grew, poverty rates in those areas rose. The assumption here is that the lower income population in these areas has the least ability or financial means to move out of the area.
In another academic research paper, author Yu Xiao of the University of Illinois at Urbana-Champaign used time‐series analysis to examine the effects of the 1993 Midwest Flood, in the paper “Local Economic Impacts of Natural Disasters”.
|Satellite imagery, courtesy of NASA, of the area around St. Louis, Missouri, in August 1991 and 1993. The deep pink scars in the 1993 image show where flood waters have drawn back to reveal the scoured land.
The flood was a significant and damaging natural disaster with major and/or record flooding occurring across 9 U.S. states. Damages totaled $15 billion with 50 deaths, resulted in hundreds of levees failing, and thousands of people evacuating over the course of some months. As damaging as the flood was, Xiao’s abstract states: “Although significant drops in personal income were observed in the year of the event, the long‐run effects seemed to be negligible.”
Both researchers’ findings showed marginal effects on the economy in the short term relative to the impact of the disaster. This may seem surprising initially, until we start to dive into the long term implications. It doesn’t always get better with time.
Getting Wind of the Long Term Impacts
So we asked a simple question earlier, “What results and trends can we expect when a natural disaster event occurs in a given region?” The quick answer we’ve uncovered from a sampling of research is that the overall economic impact is negligible. However, household incomes and home values take a dive, compounded with an uneven distribution of the population and increased poverty rate in the areas affected.
So, how well do the wounds heal over time? This is where the answer gets a little more cloudy.
A lot of debate centers on how to determine the effects of a natural disaster over a long time frame. The more time you incorporate into a study, the more variables you have to account for and the assumptions may get a little riskier. Thus, the study is difficult to accurately conduct. Plus, every disaster event is unique and in a different region. Some regions are better equipped with policy in place to aid recovery or substantial industry diversification that can withstand certain disaster events.
The Victoria University of Wellington (New Zealand) professors Ilan Noy and William duPont IV studied this topic specifically in their published research, “The Long-Term Consequences of Natural Disasters — A Summary of the Literature”. In their conclusion, the authors address the discrepancy of long term natural disaster impact results by stating:
“...disagreement most likely arises because post-disaster experiences are different in different cases, and are probably affected by the nature of hazard, the nature of exposure and vulnerability, and by post-disaster policy decisions at the levels of the households, the local authorities, national governments, and maybe even the international community’s ability and willingness to assist in recovery.”
Noy and duPont IV go on to add a very key finding that, “several case studies document long-term declines in the economic fortunes of areas that experience catastrophic events.” If we revisit the previous 1993 Midwest Flood study, Xiao also discovered, “the flood's negative impacts on agriculture were significant and long lasting to some Midwestern communities.”
Aha! We have results! So pretty much, natural disasters don’t cause much economic chaos but do create long term negative impacts, right? As I continue to learn in my study of economics, the answer is never that cut and dry.
Assessing the Damage
Regardless of the numbers and findings for any given event, there are a variety of factors to consider when studying natural disasters, but I’ll just highlight a few of what I believe are the key considerations.
When looking at total output for an affected region, the influx of government support like FEMA funding or increased GDP through reconstruction efforts can give the impression that natural disasters would be a source of economic growth. This is likely why the prior referenced studies found very negligible economic effects as a whole in the disaster areas.
Additionally, for studies using input-output models, keep in mind the data is a snapshot of the year, but doesn’t account for the fluctuation within that year. So, if $50 million worth of property or buildings were destroyed, but $50 million worth of construction was produced to restore an area later within that year, the data may show a $50 million GDP increase instead of a net of 0, after you factor in the destruction. Input-output models can model net effects, however, it’s up to the analyst conducting the study to account for this activity in his or her assumptions and model the loss and new growth afterwards.
But fear not! Research has been conducted in academia on a more precise model to measure disaster impacts. For example, two professors at the University of Illinois published a paper last year entitled, “The Challenge of Estimating the Impact of Disasters: Many Approaches, Many Limitations and a Compromise”, where they present a Generalized Dynamic Input-Output framework (GDIO) to act as a compromise between computable general equilibrium (CGE) and traditional input-output models.
If you’ve got a solid methodology and approach, but are looking to add to your research arsenal, the Federal Emergency Management Agency (FEMA) offers a free tool, HAZUS, to help with mitigation and research on the physical damage, economic loss, and social impacts of natural disasters.
|Users can download the Hazus software for free from the FEMA Flood Map Service Center (MSC).
Because of the existing limitations of studying natural disaster impacts, perhaps an alternative method would be to explore the tourism impacts or agricultural industry data for a region impacted by a natural disaster to gain a better grasp on the community impacts of a natural disaster. This can be seen in the 1993 Midwest Flood example where in the long-run, the region experienced negligible effects as a result of the flood, however agriculture experienced longer term negative impacts.
As is often the case, the assumptions and methodology are always subject to the researcher’s goal and the area of interest, like specific policy change or increased awareness of an industry or industry cluster’s needs and so on.
That being said, keep your disaster supply kit at the ready but know that getting a good grip on the economic ramifications of the aftermath of a natural disaster is a slippery undertaking. But with the right data, framework for your analysis, adopting multiple perspectives on the problem, and being open to results which describe a range of value rather than a specific number you can, in the very least, approach a clearer understanding of a post-disaster economy rather than fly blind. And no matter what, make sure you’ve got fresh batteries in your flashlight this storm season!