The Impact of Tropical Storms on Households: Evidence from Panel Data on Consumption
| Author | Nekeisha Spencer,Michael Henry,Eric Strobl |
| DOI | http://doi.org/10.1111/obes.12328 |
| Published date | 01 February 2020 |
| Date | 01 February 2020 |
1
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 82, 1 (2020) 0305–9049
doi: 10.1111/obes.12328
The Impact of Tropical Storms on Households:
Evidence from Panel Data on Consumption*
Michael Henry†Nekeisha Spencer‡Eric Strobl§
†Department of Economics, University of Birmingham, Birmingham, UK
(e-mail: m.henry.1@bham.ac.uk)
‡University of the West Indies, Kingston, Jamaica (e-mail: nspence3@binghamton.edu)
§Department of Economics and Oeschger Centre for Climate Change Research, University
of Bern, Bern, Switzerland (e-mail: eastrobl13@gmail.com)
Abstract
This paper investigates the impact of tropical storms on Jamaican household consumption.
We build a panel data set that follows individual households over time thus enabling us to
take account of time invariant household and location unobservables that could be corre-
lated with mean tropical storm exposure. Our results show that while the averagedamaging
hurricane reduces per capita consumption by approximately 1.1%, more destructive events
can cause losses multiple times this amount. There are, however, heterogeneous impacts
across households, where only those that live in buildings with less wind resistant walls
are affected. Additionally, we find that households are able to partially buffer the negative
impact on consumption through remittances and savings, as well as by shifting funds away
from non-regular expenditures. Again, households differ in the nature of this buffering
according to the wind resistance of their buildings.
I. Introduction
Tropical storms cause considerable economic losses across the world. For example, esti-
mates on their monetary damages suggest that losses have amounted to over $33 billion
since 2000.1Worryingly, developing countries are those disproportionately affected by
these extreme weather events (Doocy et al., 2013). Moreover, developing countries are
unlikely to have the social security safety nets that are typically available in developed
countries after a natural disaster; see Deryugina (2017). A growing literature has now
investigated what the aggregate wealth implications of tropical storms are, although the
conclusions have been rather mixed; see Noy and duPont (2016) and Klomp and Valckx
JEL Classification numbers: I3, Q54, Q56.
*Weacknowledge the Derek Gordon Data Bank (DGDB) at the Sir Arthur Lewis Institute of Social and Economic
Studies, University of the West Indies, Mona for providing the JSLC data. We also wish to recognize the depositor
of the data to DGDB, Statistical Institute of Jamaica.
1Source: Emergency Events Database (EM-DAT)
2Bulletin
(2014) for reviews.2However, aggregate data may be masking considerable heterogeneity
of the impact of these storms across the various actors within an economy. As a matter of
fact, in developing countries, it is often households that are vulnerable and of the greatest
concern to policymakers due to their inability to cope with negative shocks. Nevertheless,
the number of studies that have explicitly investigated the impact of tropical storms, or
even natural disasters more generally, on households have been few and their conclusions
have varied widely.
A telling insight into the current state of the literature of natural disasters on households
is the recent meta-analysis by Karim and Noy (2014). In their review of 38 papers on the
topic, the authors conclude that while there is a general tendency to find a negative impact
on household welfare,3this impact appears to be only short term. Importantly, however,
even for those studies that found clear evidence of consumption reduction, the actual
quantitative extent of these losses differs widely, thus making it difficult to evaluate the
urgencyof explicitly employing policies to help households deal with these negative shocks.
Even if one explicitly focuses on tropical cyclones, rather than natural disasters in gen-
eral, it is difficult to draw any conclusions with regardto the impact, at least quantitatively,
on household consumption. For example, Thomas et al. (2010) and Arouri et al. (2015)
find that the impact of storms resulted in Vietnamese households experiencing a 52% and
1.5% reduction in consumption with the use of cross-sectional and commune level panel
data, respectively. In contrast, provincial panel data for the Philippines show a reduction
between 5.9% and 7.1% in consumption (Anttila-Hughes and Hsiang, 2012), while Cen-
tral American countries experience a decline ranging between 2% and 4% (Ishizawa and
Miranda, 2018). Baez et al.’s (2015) use of panel data also show a fall in the consumption
of Guatemalan households by 12.6%. Thus, these storm-focused studies demonstrate a
large disparity in tropical storm impact, making it difficult to draw definitive conclusions
on how household consumption is affected.4
There could of course be many reasons as to why there is such an array of different
results on the household welfare impact of natural disasters in general, and tropical cyclones
in particular. Perhaps the most obvious, is the difference in the measurement of tropical
cyclone destruction. Earlier studies tended to use storm incidence or wind speed categories
(see Thomas et al., 2010; Arouri et al., 2015). In contrast, more recent papers modelled
local damage explicitly with physicalwind field models; see, for instance, Strobl (2012) and
Ishizawa and Miranda (2018). Importantly, capturing the appropriate functional form of
the damage function will reduce measurement error and hence the possibility of attenuation
bias.
Another potentially important reason is that, as of date, most studies on the impact of
tropical storms on household consumption have been restricted to using cross-sectional
2For instance, Smith (2008), Strobl (2011, 2012), and Elliott, Strobland Sun (2015) find shor t-term effects, while
the unpublished study by Hsiang and Jina (2014) peculiarly showsevidence of a long-term impact.
3Note that this literature has used an array of different measures for measuring welfare, including consumption
(Kurosaki, 2010; Thomas et al., 2010; Anttila-Hughes and Hsiang, 2012; Karim and Noy, 2014;Arouri, Nguyen and
Youssef, 2015), income (Karim and Noy, 2014;Arouri et al., 2015; Ishizawa and Miranda, 2018) and assets (Carter
et al., 2007).
4Note that the discrepancy between long-versus short-term impacts is also prevalent in studies for developed
countries. For instance, Deryugina, Kawano and Levitt (2014) find a long-term impact of Hurricane Katrina on its
victims, while Shaugnessy, White and Brendler (2010) found that the same storm decreased income inequality.
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
Get this document and AI-powered insights with a free trial of vLex and Vincent AI
Get Started for FreeStart Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant
-
Access comprehensive legal content with no limitations across vLex's unparalleled global legal database
-
Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength
-
Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities
-
Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting
