Natural disasters: What are the economic consequences of natural disasters for households?

Natural disasters

What are the economic consequences of natural disasters for households?

Bachelor Thesis

Author: Adélie Cleenewerck Supervisor: Thomas Giebe

Abstract

Climate change is an important subject nowadays and climate change leads to more natural disasters. This essay is a large literature study on Asian, American, European, Oceanian and African countries about the economic consequences for households as a result of natural disasters and the coping mechanisms used by households, as well as governments and institutions. It also provides information about natural disasters, such as natural disasters that have the worst consequences, people that are highly affected by disasters and places in the world where disasters happen the most. The aim of this study is to learn more about

environmental disasters and prepare better for future disasters. The results show consequences on welfare (income, assets, poverty), the labour market, migration and inequality. And the coping strategies found are post-disaster sources (help from family and relatives, public and private transfers, borrowing, credits, savings, insurance), decrease in expenditures, changes in consumption, selling assets, changes in the labour market, help from communities and other ways to cope. Governments and institutions also help households in the aftermath of natural disasters. Overall, we conclude that natural disasters lead to important economic impacts for people, and households react by using different coping mechanisms to recover.

Key words

Natural disasters, households, coping mechanisms

Acknowledgments

I would like to thank my supervisor Thomas Giebe, my examiner Mats Hammarstedt and my

opponents Sandra Morad and Carina Tran for their comments to improve my thesis.

Table of contents

1 Introduction 1

2 Methodological framework 3

3 Background theory 4

4 Economic consequences of natural disasters for households 4

4.1 Welfare of households 4

4.1.1 Income 4

4.1.2 Assets 6

4.1.3 Poverty 7

4.1.4 Welfare 8

4.2 Labour market consequences 9

4.3 Migration 10

4.4 Inequality 14

5 The worst effects of natural disasters 14

5.1 Natural disasters with the worst consequences 14

5.2 People the worst affected by natural disasters 15

5.2.1 Poor households 15

5.2.2 Women 16

5.3 Places where natural disasters happen the most 17

6 Coping mechanisms after natural disasters 17

6.1 Coping mechanisms of households 17

6.1.1 Post-disaster sources 18

6.1.2 Decrease in expenditures 20

6.1.3 Changes in consumption 20

6.1.4 Assets 22

6.1.5 Reaction in the labour market 22

6.1.6 Communities 23

6.1.7 Other examples of coping mechanisms 23

6.2 Coping mechanisms of governments and institutions 24

7 Conclusion 25

8 References 27

1 Introduction

According to the scientific community, climate change may worsen certain natural disasters and it is likely that damages after natural disasters rise in the future (UNDP, 2008). In recent years, many natural disasters have hit developed and developing countries. For example, there were cyclones such as Tauktae in India (May 2021) and Seroja in Indonesia and East Timor (April 2021), there was also an eruption of La Soufrière volcano in St. Vincent (April 2021).

Research question: What are the economic consequences of natural disasters for households?

This problem is important and relevant because more and more environmental disasters have happened over the recent years because of climate change, and households as well as

governments and institutions need to be prepared to face more frequent natural disasters. Such events have always happened in the past, but climate change and the rise of temperatures might result in more or worse natural disasters. Natural disasters cause humanitarian and economic costs. There exist many studies about macroeconomic consequences, but here we will see that households also suffer from economic consequences. It can be interesting to see the economic impacts of natural disasters for households. There have been a lot of previous studies on natural disasters, however there are fewer empirical studies on the economic impact of natural disasters for households. In general, studies focused on a specific country and on a specific economic consequence, and so this essay will contribute to the literature by summarising different economic consequences in several countries in the world and showing what mechanisms households use to cope after natural disasters.

Asia is the continent with the most economic damages caused by natural disasters, 60.6

percent of the total economic losses in the aftermath of natural disasters came from Asia in

2019 (Statista, 2020). As a consequence, this essay will look at many studies about Asian

countries, such as Vietnam, the Philippines, China, Bangladesh, Indonesia, Japan, Sri Lanka,

India and Thailand. First, there exist some studies about natural disasters in Vietnam, such as

Arouri et al. (2015), Bui et al. (2014), Trinh et al. (2021) and Thomas et al. (2010). For the

Philippines, we will see a study by Israel and Briones (2014), studies about typhoons (Anttila-

Hughes and Hsiang, 2013; Skoufias et al., 2019), and more precisely, we will look at the

typhoons Ondoy and Pepeng in 2009 (GOP, 2009a, 2000b) and Typhoon Milenyo in 2006 (Sakai et al., 2017; Sawada et al., 2009). About China, this thesis will analyse the Wenchuan earthquake in 2008 (Feng et al., 2016) and a study by Pei (2017) in historical China. For Bangladesh, we will see studies about natural disasters written by Islam (2017) in three Northern Districts of Bangladesh, Shahabuddin and Ali (2006) and Ishtiaque and Nazem (2016) in Dhaka city. We will also examine floods in Bangladesh such as Karim (2018), Gray and Mueller (2012) and Paul and Routray (2010a) in Bannabari and Suvagacha villages. More precisely, we will look at the 1998 floods (Mueller and Quisumbing, 2011; Del Ninno et al., 2003; Shah, 1999). This thesis will also study cyclones with Paul and Routray (2010b) in coastal Bangladesh, Alam et al. (2018) about cyclones Aila, Sidr and Mohaseen in the Kuakata coastal belt of Patuakhali Bangladesh, and also Cyclone Aila in 2009 (Abdullah et al., 2016; Mottaleb et al., 2013). Finally, we will examine the tornado that hit the north of Bangladesh in 2004 (Paul, 2005). About Indonesia, this essay will look at a study by Kirchberger (2017) about the earthquake that hit Yogyakarta in 2006 and Tse (2011). For Japan, we will analyse the Great Hanshin‐Awaji earthquake in 1995 (Sawada and

Shimizutani, 2008) and the Great East earthquake and tsunami in 2011 (Venn, 2012).

Concerning Sri Lanka, we will see a study written by Keerthiratne and Tol (2018). As for India, there is a study about extreme precipitation events in Mumbai, Chennai, and Puri district (Patankar, 2019). Finally, Francisco et al. (2011) studied natural disasters in Southeast Asian countries (i.e., China, the Philippines, Vietnam, Thailand and Indonesia).

Then, 28.9 percent of the total economic losses as a result of natural disasters came from America in 2019 (Statista, 2020). We will study some American countries, such as Mexico, Nicaragua, Honduras, El Salvador, the United States and Chile. In Mexico, this essay will be about studies such as Saldaña-Zorrilla (2008) and Saldaña‐Zorrilla and Sandberg (2009). For Nicaragua, we will look at a study made by Loebach (2016). Van den Berg and Burger (2008) and Carter et al. (2007) studied Hurricane Mitch (1998) in Nicaragua and Honduras. Then, Baez and Santos (2008) examined the two earthquakes that hit El Salvador in 2001. Besides, there are some studies about natural disasters in the United States, such as Zeenat Fouzia et al.

(2019). About hurricanes, we will see Hurricane Katrina in 2005 (Masozera et al., 2007;

Venn, 2012), Hurricanes Katrina and Rita in 2005 (Do Yun and Waldorf, 2016) and

Hurricane Andrew (1992) in Dade County (Florida) (Smith and McCarty, 1996). Finally, there exist studies about the earthquake and tsunami in Chile (2010), such as Venn (2012), Moreno (2018) and Martínez et al. (2020).

About Europe, 4.2 percent of the total economic losses in the aftermath of natural disasters came from Europe in 2019 (Statista, 2020). This thesis will look at two earthquakes that happened in Italy: the Messina-Reggio Calabria earthquake in 1908 (Spitzer et al., 2020), and L’Aquila’s earthquake in 2009 (Ambrosetti and Petrillo, 2016).

Moreover, 3.9 percent of the total economic losses as a result of natural disasters came from Oceania in 2019 (Statista, 2020). This essay will analyse the Canterbury earthquakes in New Zealand (2010/2011) (Venn, 2012).

And finally, 2.4 percent of the total economic losses in the aftermath of natural disasters came from Africa in 2019 (Statista, 2020). In Africa, we will see Nigeria (Adeagbo et al., 2016), Ethiopia (Carter et al., 2007) and Uganda (Helgeson et al., 2013).

The paper is structured in five sections. Section 2 contains the methodological framework.

Section 3 presents the background theory. Section 4 shows the economic consequences for households in the aftermath of natural disasters. Section 5 presents the worst effects of natural disasters. Section 6 describes how households, and governments and institutions cope with disasters. Section 7 concludes.

2 Methodological framework

This essay will be a large literature study. First, we will talk about the economic

consequences of natural disasters for households: welfare of households (income, assets and

poverty), labour market consequences, migration and inequality. Then, we will see disasters

that have the worst consequences, people that are the worst affected and the parts of the world

where they happen the most. Finally, we will look at the coping mechanisms of households

and analyse how countries, governments, institutions cope with natural disasters, in order to

know how households can prepare and learn to live with natural disasters.

3 Background theory

Natural disasters have short-term and long-term economic implications. The economic effects are the fact that households experience direct losses (e.g., destruction of infrastructures, houses and factories, damages to assets, flooding, etc.) and indirect losses (e.g., losses due to higher prices of food and other necessities, decline in income, shortages of food, water, and fuel, disruption of services, loss of work days and jobs, no tourists coming, no harvest, etc.) after a natural disaster. These effects are likely to make the income of households lower, which will make poor households poorer and then poverty may increase. As a consequence, this will reduce their overall household welfare. As a result, households respond (e.g., adjustment in expenditures, consumption, assets, borrowing, change of jobs, migration, etc.) and governments react (e.g., financial help, etc.).

Figure 1. Consequences of natural disasters

4 Economic consequences of natural disasters for households

4.1 Welfare of households

This section will investigate the impacts of natural disasters on various aspects of welfare, such as income, assets and poverty.

4.1.1 Income

This section starts with results on household income.

Some studies in Vietnam found that the income of households living in areas exposed to natural disasters are much lower than the average income (Thomas et al., 2010; Bui et al., 2014). Most studies found declining incomes after natural disasters. Arouri et al. (2015) showed that per capita income of rural households was reduced by natural disasters in

Vietnam. Bui et al. (2014) discovered that an exposure to environmental disasters in Vietnam caused a significant decrease in income, of 7 percent in the average income of households.

For the Philippines, households that were affected by a natural disaster in Pasay city in the last twelve months were more likely to have lower per capita incomes (by 6.7 percent) than households that were not affected (Israel and Briones, 2014). Anttila-Hughes and Hsiang (2013) discovered that the income of an average Filipino household was 6.6 percent lower in an average year one year after typhoons. According to the GOP (2009a, 2000b), income decreased by 12 percent of the average household income in the regions affected by the typhoons Ondoy and Pepeng. Sawada et al. (2009) discovered that income decreased for 54 percent of farmers and 27 percent of landless after Typhoon Milenyo. Then, after the Wenchuan earthquake in China, income of households declined by 14 percent (Feng et al., 2016). In Bangladesh, household income decreased after the 1998 flood. For instance, an imminent decrease in wages has been noticed in the short run, according to Mueller and Quisumbing (2011), Del Ninno et al. (2003) found that the average monthly earnings in July–

October 1998 were 46 percent lower than those in July-October 1997 for daily labourers, and Shah (1999) discovered that real wages fell. Paul and Routray (2010a) showed that there was a loss of earnings in two villages of Bangladesh (Bannabari and Suvagacha). Then, natural disasters in Mexico had a negative effect on incomes of households according to Saldaña‐

Zorrilla and Sandberg (2009). For Nicaragua, Van den Berg and Burger (2008) found that households dealt with income shocks (for most people a decline in income) after Hurricane Mitch. Baez and Santos (2008) discovered that the earthquakes in El Salvador declined the income per capita by one-third, compared to the pre-shock average. Besides, Hurricane Katrina in the United States had severe impacts on income of households (Masozera et al., 2007). People that had to leave their homes in the disaster area of the hurricanes Katrina and Rita had a lower income than if they could have stayed (Do Yun and Waldorf, 2016).

Moreover, the monthly wages of households declined by 6.4 percent after the Bío-Bío

earthquake and tsunami in Chile, this could be explained by the fact that the natural disaster caused damages to the workplace (Martínez et al., 2020). Most households answering a survey in Nigeria (64.6 percent) affirmed that natural disasters had a negative impact on their income (Adeagbo et al., 2016). Finally for Ethiopia, households faced losses in incomes in the long run because of the three-year drought (Carter et al., 2007).

Concerning agriculture, Karim (2018) found evidence of negative effects on income (crop and non-crop) of agricultural workers in the short run after recurrent flooding in Bangladesh.

Ishtiaque and Nazem (2016) showed that a flood in Bangladesh caused important damage to crops and as a consequence, income from agriculture experienced a sharp decline. Cyclone Aila in Bangladesh caused an important decline in rice income for households (Mottaleb et al., 2013). Then, Trinh et al. (2021) discovered that natural disasters in Vietnam had a negative impact on crop revenue. Carter et al. (2007) discovered that Hurricane Mitch in Honduras caused a loss of 30-40 percent of poor rural households’ crop income, shortly after the event. Some studies showed that agricultural households experienced a higher impact on the decline in income in the short run, such as after Typhoon Milenyo in the Rural Philippines (Sakai et al., 2017) and after the 1998 flood in Bangladesh. Some agricultural workers went towards non-agricultural employment to have a lower decrease in wages (Mueller and Quisumbing, 2011).

In contrast, household income can go up, as found by Kirchberger (2017). The author studied the impacts of the earthquake that hit Yogyakarta in Indonesia and discovered that workers in the agricultural sector, who had the lowest wages, had faster wage growth. It had a positive impact on wage growth for these workers. This can mostly be explained by the fact by the shift of workers out of the agricultural sector into the construction sector, thus making the marginal product of labour increase in agriculture.

4.1.2 Assets

Natural disasters have consequences on assets.

In general, households lost assets in the aftermath of environmental disasters. For example,

there was a reduction in asset holdings after natural disasters in Vietnam (Thomas et al.,

2010). Households in the Philippines lost physical assets after typhoons (Anttila-Hughes and

Hsiang, 2013). It was also the case in Bangladesh, for example, the 1998 floods damaged private assets (Shah, 1999). According to Paul and Routray (2010a), there was a loss of assets in two villages of Bangladesh (Bannabari and Suvagacha). For Indonesia, earthquakes

lowered non-business assets by 69 percent and they also caused damages to housing assets (Tse, 2011). The Great Hanshin-Awaji earthquake (Japan) caused damage to assets according to 87 percent of households (Sawada and Shimizutani, 2008). In Chennai (India), households were affected by damage to assets after floods (Patankar, 2019). After Hurricane Mitch, households dealt with asset shocks, such as a loss of land or housing, in Nicaragua (Van den Berg and Burger, 2008), and some households lost almost all their productive assets in Honduras and households that had lower wealth lost 15–20 percent of their productive assets, shortly after the hurricane, which had consequences on earnings and livelihood (Carter et al., 2007). Then, about two thirds of households experienced asset losses after the earthquakes in El Salvador (Baez and Santos, 2008). Moreover, there was a negative impact on household assets after natural disasters in Nigeria, 41 percent of respondents of a survey said that they experienced severe negative effects on assets (Adeagbo et al., 2016).

However, Tse (2011) found that eruptions in Indonesia increased farm assets by 55 percent (this may be because lava ash boosts soil fertility) and housing assets (this can be explained by the implementation of relief money to rebuild houses).

4.1.3 Poverty

Environmental disasters impact poverty.

This study found that environmental disasters usually increase poverty. For example in Vietnam, there was a correlation between poverty rates and disasters (Arouri et al., 2015).

According to Bui et al. (2014), natural disasters in Vietnam may make poverty worse, with a

degradation in poverty for the affected households. In the Philippines, floods and typhoons in

Pasay City had a negative and significant effect on poverty among poor households (Israel

and Briones, 2014). Besides, Shahabuddin and Ali (2006) found that the poverty situation is

worse in the ecologically vulnerable zones than in the favourable zones for rural households

in the disaster-prone areas of Bangladesh and more than two-third of the households of the

vulnerable zones are considered as poor. Then, Carter et al. (2007) discovered that poverty

increased 5.5 percentage points immediately after Hurricane Mitch in Honduras and that there was a poverty trap. According to Adeagbo et al. (2016), rural residents suffered more from impoverishment in Nigeria after natural disasters.

However, some studies found that there was not a large impact on poverty after natural disasters. For instance, there was only a small effect on the poverty gap (0.3 percent in 2009 and 0.4 percent in 2010) and severity of poverty (0.2 percent in 2009 and 0.3 percent in 2010) after the typhoons Ondoy and Pepeng in the Philippines (GOP, 2009a, 2000b). And,

according to Van den Berg and Burger (2008), Hurricane Mitch did not cause a significant change in the overall poverty profile in Nicaragua in 1999. Finally, the earthquakes in El Salvador had only a minor impact on poverty, however they had an impact on the standards of living of low-middle-income and middle-income rural households (Baez and Santos, 2008).

4.1.4 Welfare

Welfare of households is impacted after natural disasters.

Household welfare usually decreased after natural disasters. In Vietnam, there was a negative effect on household welfare, according to Arouri et al. (2015). Thomas et al. (2010) showed that there was a welfare loss in areas often exposed to droughts and hurricanes in Vietnam.

Riverine floods decreased welfare of households by 23 percent (by 5.8 percent immediately and by 17.2 percent in the following year). Hurricane force winds were the largest damage, they caused damage to household welfare immediately and that lasted for many years.

Hurricanes led to a 52 percent immediate loss of welfare in cities in Vietnam with more than 500,000 inhabitants.

However, there might be a positive impact on welfare. For instance in Vietnam, welfare increased in areas frequently exposed to excess rain (Thomas et al., 2010). Then, after Typhoon Milenyo in the rural Philippines, the price of fish experienced a sharp decline because the supply of fish increased, this can be explained by the fact that many cultured fish from a lake nearby were released after the typhoon damaged fish pond facilities. As a

consequence, rich people had positive net gains in welfare due to the decrease in fish price. It

can be surprising but the welfare effects of the changes in price were mostly positive (Sakai et

al., 2017). And according to Shahabuddin and Ali (2006), some rural households in the

disaster-prone areas of Bangladesh could sustain or even improve their economic position, as before the natural disaster happened. This is thanks to the relationship households had, also because they had a smaller size, a better health and migratory tendency, they changed jobs, and they were motivated.

4.2 Labour market consequences

The labour market experiences consequences in the aftermath of natural disasters.

Natural disasters usually have negative impacts on the labour market. According to Venn (2012), conditions in the labour market were worse after natural disasters in OECD countries, such as Japan, the United States, Chile and New Zealand.

First, natural disasters affect employment. Del Ninno et al. (2003) found that the 1998 flood in Bangladesh decreased employment opportunities for daily labourers. Then, employment might decrease, such as in the regions of the earthquakes in New Zealand (by 14 percent in Christchurch and by 8.3 percent in the Canterbury region between December 2010 and December 2011) and after Hurricane Katrina in the United States with a decrease in employment in the regions the most concerned, by 35 percent in New Orleans and by 15 percent in Gulfport-Biloxi (Venn, 2012). Zeenat Fouzia et al. (2019) found that there are different effects on employment depending on the climate-related disasters in the United States. For instance, employment usually increases in the year following wildfire, hurricane and heatwaves, and it often decreases the year after flooding and severe storms. According to Martínez et al. (2020), there was a negative effect on the quantity and quality of employment in the months following the Bío-Bío earthquake and tsunami in Chile (in the short run).

However, signs of recovery in the Chilean labour market have been noticed in the long term, for instance there were positive effects on the probability of being employed and on the probability of participation in the labour force, this may be because households who lost their jobs after the earthquake and tsunami came back to the affected areas.

Some people might lose their jobs. Some firms closed because of damages, and the staff and

customers missing, this put some jobs at risk. After a natural disaster, some workers were

more affected by the losses of jobs than others. For example, 90 000 workers lost their jobs in

Chile. After Hurricane Katrina, there were less workers in the regions that were the worst affected. As a consequence, the staff that stayed had a reduced workload (Venn, 2012).

Some households lost days of work. For example, people working in paid and self-

employment in Nigeria said that they lost days of work, on average 54.5 days, this led to a decrease in economic activity (Adeagbo et al., 2016).

Unemployment may increase. It was the case after the 1998 floods in Bangladesh (Shah, 1999) and in the most-affected areas caused by the Great East Japan earthquake and tsunami in Japan (Venn, 2012).

Natural disasters may have a negative impact on the level of productivity/production at work of households. It was the case in Nigeria, most people answering a survey in Nigeria (50.9 percent) affirmed that natural disasters had a negative impact on their level of

productivity/production at work (Adeagbo et al., 2016).

However, Kirchberger (2017) found that labour market outcomes for individuals who were living in the affected regions of the earthquake that hit Yogyakarta in Indonesia were quite resilient to the natural disaster.

4.3 Migration

One of the primary responses to natural disasters is usually migration. Natural disasters lead to small-scale and large-scale migration of people, to get away from affected areas. People lose their livelihoods and income opportunities after natural disasters, and this causes out- migration. Households migrate to non-affected areas because they are looking to be employed and also because they fear that there will be more natural disasters in the future. After natural disasters, households of rural areas of developing countries tend to migrate to large cities.

According to previous studies, in developing countries, migration seems to be one of the

coping mechanisms for natural disasters (Drabo and Mbaye, 2015). Krishnamurthy (2012)

suggested that natural disasters are not the cause of migration, but they may intensify the

economic conditions of households and then households decide to migrate. Researchers

disagree about the consequences of natural disasters on migration, some think that they

increase migration, some that they decrease migration, or that they do not have an impact on

migration. Forced migration will depend on the resilience relative to damage and the decline in income.

According to some studies, migration increased after environmental disasters. Drabo and Mbaye (2015) studied international migration from developing countries to developed countries after natural disasters and found that natural disasters were positively related with migration rates. Besides, the impact of natural disasters is different according to the education level, it may worsen the brain drain in developing countries because of the migration of only highly skilled people. Furthermore, the effects of natural disasters on migration will depend on the location of countries and depend on the type of disaster. Trinh et al. (2021) discovered that if natural disasters are more severe, then there will be a higher probability of migration in Vietnam (a 10 percent increase in the severity of natural disasters will lead to an increase in the probability of a household sending a migrant by 0.2 points) and the loss in income

resulted in a higher probability of migration. Pei (2017) found that migration of agriculturists after natural disasters in historical China was almost instantaneous and only effective in the short run. In general, people can recover from the short run effects of natural disasters. After natural disasters, agriculturists need to move until the impacts of disasters have disappeared, then migration will stop. The impacts of natural disasters cannot last in the long run, however, population pressure can push migration in the long run. For Bangladesh, Islam (2017) looked at 28 chars (a char is a riverine island in Bengali) and found that char people experienced consequences because of natural disasters and economic vulnerabilities. This increased their decision to migrate from one char to another. Natural disasters make poor people in

vulnerable locations think about migrating. The author found that char people tend to migrate after natural disasters. Shahabuddin and Ali (2006) discovered that households usually migrated, in the disaster-prone-areas in Bangladesh. According to Ishtiaque and Nazem (2016), 18 percent of rural-urban migrants were because of natural disasters in Dhaka city.

Most of these migrants moved out because they experienced a sharp decline in income

immediately after the natural disaster. Gray and Mueller (2012) showed that natural disasters

did not have positive impacts on overall mobility in rural Bangladesh. Flooding led to short

run movement of population, but they caused little long-run displacement. Also, households

living in affected areas had a higher probability to move. Paul and Routray (2010a) found that

some households migrated temporarily in two villages of Bangladesh, 54.3 percent of households in Bannabari and 94.9 percent of households in Suvagacha. Paul and Routray (2010b) discovered that households migrated in the short run after cyclones and induced surges in coastal Bangladesh. Then, the two prefectures that were the most affected by the Great East earthquake and tsunami in Japan encountered net outflows of households in the four months following the disaster and they still did not recover 18 months later (Venn, 2012). Francisco et al. (2011) found that many households evacuated and moved to safer places after natural disasters in Southeast Asian countries (i.e., China, the Philippines, Vietnam, Thailand and Indonesia), this was the most important behavioural adaptation

measure in the five countries. Moreover, in Mexico, farmers planned to emigrate immediately after natural disasters (Saldaña-Zorrilla, 2008). Saldaña‐Zorrilla and Sandberg (2009) found that regions that are often hit by natural disasters tend to have higher emigration rates. For instance, if the disaster frequency increased in a region by 10 percent, then this will raise emigration rates by 1 percent. This is particularly the case for higher educated people because they know better about emigration. Also, workers who experienced a significant decrease in wages were most likely to out-migrate. Do Yun and Waldorf (2016) found that some people were forced to move in the aftermath of Hurricanes Katrina and Rita. Smith and McCarty (1996) discovered that Hurricane Andrew, that hit the United States, drove more than 353,000 people from their homes temporarily. This corresponds to one of the largest population moves in response to a natural disaster in the United States. And about 40,000 people left

permanently. Approximately 17 percent of the survey respondents living in Dade County in August 1992 were forced to leave their homes after the hurricane, migration after the

hurricane was mostly in the short run. About 67 percent of people that had moved, returned to their homes by August 1994. Other moves lasted for many months, and some were

permanent. According to Martínez et al. (2020), the probability of emigration increased after the Bío-Bío earthquake and tsunami in Chile, in particular for workers. Finally, in Nigeria, about 40 percent of respondents of a survey indicated that natural disasters led to

displacement, about half for rural respondents and 26.9 percent for urban respondents

(Adeagbo et al., 2016).

Natural disasters sometimes have only a small impact on migration. For example, Spitzer et al. (2020) found no evidence that the Messina-Reggio Calabria earthquake in Italy had, on aggregate, a large impact on emigration, this was a negligible and short-lived impact. But there was a positive effect on emigration in districts where there were many agricultural workers. However, this does not mean that the earthquake had no effect on emigration. The authors found that if people were not really attached to the land, then they were more likely to migrate. Ambrosetti and Petrillo (2016) showed that there was no massive population

migration after L’Aquila’s earthquake in Italy. Still, they found an increase of out-flows from L’Aquila to other provinces, in particular to the closest regions (Latium and Campania), after the earthquake. In addition to the earthquake, L’Aquila province had already economic difficulties, this worsened the problems and the recovery process took a long time. And so migration could be a solution for some households.

Natural disasters sometimes might not have an impact on migration at all. Paul (2005) discovered that households suffered from damages after the tornado in the north of

Bangladesh and it led to important damages to field crops and other property. But he did not find evidence of out-migration after the tornado, all 291 respondents said they did not migrate after the tornado. It might be thanks to the disaster relief that households received after the event and so they did not feel the need to leave their homes. Another reason might be the fact that jobs were created in the affected villages. Or it could be because the disaster did not affect an important area, for example floods or cyclones concern a larger area and lead to out- migration. Finally, households might also have an attachment to their homes. The author also found that emergency aid may help to recover from damages, even if this is not valid for all countries. Keerthiratne and Tol (2018) did not find internal migration as a result of

environmental disasters in Sri Lanka. Finally, Loebach (2016) did not find evidence on

livelihood international migration of individuals after Hurricane Mitch in Nicaragua, but there was an increase in the migrant selectivity. Also, Mitch caused a decline in the likelihood of migration of small business owners.

Migration can actually decrease as a result of a natural disaster. For instance, Tse (2011)

found that natural disasters in Indonesia led to a significant decline in migration rates, so

natural disasters may actually decrease migration and households move out less. Disasters

destroyed assets of households and this discouraged migration. Moreover, natural disasters may increase the marginal product of labour in some sectors, and this causes a reduction in migration. There is evidence that shows that if a household is smaller and has lower earnings, then migration is less likely to happen. Finally, Baez and Santos (2008) discovered that the two earthquakes that hit El Salvador led to a decline in migration.

4.4 Inequality

Environmental disasters might create or worsen inequality between households.

Natural disasters are expected to increase inequality. According to Bui et al. (2014),

environmental disasters rose income inequality among households in Vietnam. It may worsen the gap between rich and poor households. Carter et al. (2007) found that there was a high level of inequality after Hurricane Mitch in Honduras.

Sometimes, natural disasters do not have impacts on inequality. For instance, the impact on income inequality was negligible (0 percent in 2009 and 2010) after the typhoons Ondoy and Pepeng in the Philippines (GOP, 2009a, 2000b). Also, Feng et al. (2016) found that there was no change in income inequality after the Wenchuan earthquake in China.

Inequality may actually decrease. For instance, income inequality decreased after

Cyclone Aila in Bangladesh in the short run (Abdullah et al., 2016) and after natural disasters in Sri Lanka, this reflects a decline in income of the rich quintiles (and not an increase in income of the poorest quintile). Rich people were more affected than poor people because their income is mostly based on activities that are non-agricultural and non-seasonal

agricultural, and environmental disasters lower non-agricultural and non-seasonal agricultural income inequality (Keerthiratne and Tol, 2018).

5 The worst effects of natural disasters

5.1 Natural disasters with the worst consequences

This thesis does not find which natural disasters lead to the worst consequences because

studies found different results.

Natural disasters in the Pacific Rim (from 1970 to 2008) that had the worst damages (in US dollars) were a storm in the United States (2005), an earthquake in Japan (1995), a flood in China (1998), an earthquake in China (2008) and an earthquake in the United States (1994) (Noy, 2012).

Shahabuddin and Ali (2006) found that the worst disaster is flood in Bangladesh, by looking at food shortages, income and poverty. As a consequence, there are more government

programs like the Vulnerable Group Development (VGD) and the Vulnerable Group Feeding (VGF) in the flood-prone zones. The VGD is a program that aims to provide food and

improve the livelihood of poor and vulnerable women after wars, famines and floods. And the VGF program aims to give food to poor households after natural disasters, such as floods and cyclones. Ishtiaque and Nazem (2016) confirmed that flood led to the most important

damages to crops in Dhaka city (Bangladesh).

According to Pelling et al. (2004), earthquakes appear to be the most expensive disasters and floods do not lead to large losses.

In the United States, tropical cyclones caused the largest damage (from 1980 to 2019), approximately $927.5 billion (CPI adjusted) (Sparks, n.d.).

5.2 People the worst affected by natural disasters

5.2.1 Poor households

Studies generally show that poor households are highly affected by natural disasters because of the reduction in income, in their asset base, in their livelihood opportunities, which make poverty worse and decrease the overall welfare of households. Fothergill and Peek

(2004), Peacock et al. (1997) and Wisner et al. (2004) suggested that poor households in

developing and developed countries were more vulnerable to natural shocks. Kim (2011)

showed that poor households were more exposed to natural disasters than non-poor

households. In Vietnam, poor households are more likely to live in communes with high

probabilities of disasters (Arouri et al., 2015). Natural disasters in Vietnam had worse effects

on poor households (Bui et al., 2014). Thomas et al. (2010) found that households the more

exposed to natural disasters tend to be poor in Vietnam. For the Philippines, there was a

significant difference between poor and non-poor households in the reduction of the

consumption of rice after Typhoon Milenyo. The price of rice rose, and this has directly affected expenditures of poor households. The impact on welfare was also different between poor and non-poor households, non-poor households experienced positive welfare impacts (significant) and poor households did not gain from the reduction in the price of fish because they do not consume that much fish (Sakai et al., 2017). In India, poor households were more vulnerable to natural disasters because they had lower incomes, housing of poor quality, and they lived in flood-prone areas. However, they faced lower damage costs because they owned less than non-poor households (Patankar, 2019). Do Yun and Waldorf (2016) found that the most affected by Hurricanes Katrina and Rita were low-income households. Finally, Carter et al. (2007) showed that there were differences in growth, meaning that poor households are more sensitive to shocks, and will be less able to recover after the drought in Ethiopia and Hurricane Mitch in Honduras.

5.2.2 Women

Women seem to be more vulnerable to natural shocks.

According to some studies (Agarwal, 1990; Begum, 1993; Walker, 1994; Fordham, 1998;

Cannon, 2002; Hutton and Haque, 2004; Ray-Bennett, 2009), the effects of natural disasters are higher for women than men. Women are considered as the worst victims and the most vulnerable. For example, Venn (2012) found that employment of women declined by 10 percent, compared to 7 percent for men after the Canterbury earthquakes in New Zealand, this can be explained by the fact that there were higher losses in employment in women-

dominated industries such as hospitality, retail, etc. Women were also more affected by the earthquake and tsunami in Chile, with 60 percent of net job losses, and the labour market participation for unskilled women was highly affected because the women-dominated industries were among the worst affected (i.e., tourism, services, etc.). After Hurricane Katrina in the United States, women usually had more difficulties to recover than men because they had lower wages and family responsibilities (Masozera et al., 2007).

The International Labour Organisation (ILO) suggested that women are more vulnerable

during natural disasters. This can be explained by the gender division of labour, i.e., the

allocation of different jobs to women and men. They also seem to have less access to

resources, and most women have to take care of their children. As a consequence, they are less capable of finding resources to recover, they have more chances to be unemployed and they do not earn an income. Women are also subject to the risk of sexual violence after a natural disaster.

5.3 Places where natural disasters happen the most

According to Pelling et al. (2004), 85 percent of households exposed to natural disasters (i.e., earthquakes, tropical cyclones, floods and droughts) live in countries with medium or low levels of development. Asian countries experience the worst economic losses, where a lot of floods happen in this region. Earthquakes happen the most in countries such as Iran,

Afghanistan, India, Turkey, Armenia and Guinea. Tropical cyclones usually take place in Bangladesh, Honduras, Nicaragua, India, the Philippines and Vietnam. There is a high vulnerability to floods in many countries, for instance in Venezuela, Somalia, Morocco and Yemen. And droughts happen the most in African countries.

Kostigen (2020) suggested that the Philippines, China, Japan and Bangladesh are the countries where there are the most natural disasters, such as storms, floods, earthquakes, volcanoes, tsunamis, wildfires and landslides.

Kim (2011) found that poor households in East Asia and Pacific are the most exposed to natural disasters, then in South Asia and Sub-Saharan Africa. However, the exposure is increasing in South Asia and Sub-Saharan Africa and decreasing in East Asia and Pacific.

6 Coping mechanisms after natural disasters

In this section, findings from the literature regarding how households and governments cope with the effects of environmental disasters will be presented. This is important because we need to learn from previous disasters in order to prepare better for the future.

6.1 Coping mechanisms of households

Coping mechanisms of households are the choices that households make to recover from

natural disasters. Typically, households’ welfare declines as shown in section 4.1.4, but there

are some exceptions where households show a certain resilience, or even increasing incomes.

Perhaps, households used coping mechanisms to recover from natural disasters.

Paul and Routray (2010a) discovered that coping strategies (changes in eating behaviour, borrowing money, disposing assets, migration, etc.) after floods in Bangladesh helped households to decrease their vulnerability. The different coping strategies showed how households reacted to floods, depending on income, education and occupation. If household income increases, access to food and water will increase, this increases the capacity to cope with floods. For education, the higher the education level, the higher the information, and the higher the capacity to cope with a disaster. As for occupation, some sectors have a more stable income, for instance the service sector, and workers have a higher ability to cope with disasters. Paul and Routray (2010b) found that households used coping strategies after

cyclones and induced surges in coastal Bangladesh and these coping mechanisms will depend on the frequency and intensity of the natural disaster, and on economic factors.

6.1.1 Post-disaster sources

First, households have access to different post-disaster sources.

Households can be helped by people they know. For instance, the coping strategies after Typhoon Milenyo in the Philippines were emergency loans from relatives, village

moneylenders and remittances (Sakai et al., 2017). Saldaña-Zorrilla (2008) found that the most important financing sources for farmers were relatives in the community and

neighbours, after natural disasters in Southern Mexico. According to Adeagbo et al. (2016), Nigerian households were helped by relatives and friends.

Households can benefit from public and private transfers. This was the case for rural

households after the changes in consumption (Wainwright and Newman, 2011) and after the Great Hanshin‐Awaji earthquake in Japan (Sawada and Shimizutani, 2008).

One of the main coping strategies is to borrow. It was the case in Bangladesh. For instance,

households usually borrowed in the disaster-prone-areas of Bangladesh (Shahabuddin and

Ali, 2006). Households also borrowed after the 1998 floods: it was the most important coping

mechanism of households, in order to purchase food and other expenses (Del Ninno et al.,

2003) and rural poor households borrowed money from relatives and lenders (Shah, 1999).

According to Paul and Routray (2010a), borrowing money was an important initiative after the floods in two villages of Bangladesh (Bannabari and Suvagacha). People also borrowed money after cyclones and induced surges in coastal Bangladesh (Paul and Routray, 2010b).

Abdullah et al. (2016) found that borrowing money was the most common coping strategy after Cyclone Aila. Then, after the Great Hanshin‐Awaji earthquake in Japan, some

households borrowed, according to Sawada and Shimizutani (2008). Households in India that could not benefit from the compensation given by the government, had to borrow from informal sources, after extreme precipitation events (Patankar, 2019). Carter et al. (2007) found that households may borrow against future earnings, in order to maintain their

consumption levels after natural disasters in Ethiopia and Honduras. Finally, poor households may also borrow after income shocks (Ghorpade, 2012).

Another coping strategy that households used is credits. For example, there was a change in the credit market after natural disasters in Vietnam (Arouri et al., 2015). Credit makes

households more resilient to shocks (Briguglio et al., 2009, Bruneau et al., 2003, Davies et al., 2013).

Besides, households used their savings. For example, households usually used their savings in the disaster-prone-areas in Bangladesh (Shahabuddin and Ali, 2006). Rural poor households spent their savings after the 1998 floods in Bangladesh (Shah, 1999). And households in India had to use their savings after extreme precipitation events if they could not benefit from the compensation given by the government (Patankar, 2019).

Finally, households can get insurance. After the Great Hanshin‐Awaji earthquake in Japan, a few households benefited from earthquake insurance (Sawada and Shimizutani, 2008).

Francisco et al. (2011) discovered that households did not buy disaster insurance in Southeast Asian countries, but some households were willing to pay for insurance, 46 percent of

Chinese were willing to pay for insurance, 37 percent of Thai, 33 percent of Filipino and 20

percent of Vietnamese. However, in India, more than 90 percent of the affected households

had no insurance at all, and for people that had one, it did not cover the whole extent of the

losses (Patankar, 2019).

6.1.2 Decrease in expenditures

Households decreased their expenditures in most cases after environmental disasters. Some studies in Vietnam (Bui et al., 2014 and Thomas et al., 2010) showed that expenditure of households living in areas exposed to natural disasters are much lower than the average expenditure. Per capita expenditure of households declined in Vietnam, of 8.3 percent according to Bui et al. (2014) and for rural households according to Arouri et al. (2015). For the Philippines, the reduction in income led to a 7 percent decrease in expenditures after typhoons (Anttila-Hughes and Hsiang, 2013). Skoufias et al. (2019) found that typhoons in the Philippines declined the total per capita expenditures, in particular for food and protein.

According to Sawada et al. (2009), there was a reduction in the expenditure on rice, by 7 percent after Typhoon Milenyo. According to Shahabuddin and Ali (2006), households usually reduced their expenditures on necessities (e.g., education, health, clothing and housing) and food consumption in the disaster-prone-areas of Bangladesh. Households also reduced their expenditures after the 1998 floods in Bangladesh (Del Ninno et al., 2003). Then, there was evidence of negative effects on expenditure of agricultural workers after recurrent flooding in Bangladesh, in the short run (Karim, 2018). Mottaleb et al. (2013) discovered that the affected households spent less on the education of their children and on health after the income loss caused by Cyclone Aila in Bangladesh. After the Great Hanshin-Awaji

earthquake in Japan, households were forced to decrease their expenditures (Sawada and Shimizutani, 2008). Finally, Helgeson et al. (2013) found that the second most frequently coping strategy used in rural Uganda was to decrease expenditures.

6.1.3 Changes in consumption

Households adjust their consumption. For instance, after Typhoon Milenyo in the Philippines,

households shifted their consumption from purchased to home-produced items (Sakai et al.,

2017) and rural households changed their consumption expenditure because of the changes in

commodity prices caused by the typhoon (Sawada et al., 2009). Paul and Routray (2010b)

discovered that households bought inexpensive foods after cyclones and induced surges in

coastal Bangladesh. After the Great Hanshin‐Awaji earthquake in Japan, a coping mechanism

of households was to change the composition of their consumption (Sawada and Shimizutani,

2008). In Nicaragua, asset-poor households adjusted their consumption after changes in income in the aftermath of Hurricane Mitch (Van den Berg and Burger, 2008).

In some countries, households reduced their consumption. For example, natural disasters in Vietnam lowered consumption of households by 7.1 percent (Bui et al., 2014). Concerning a drought, Vietnamese households with no irrigated plots experienced a reduction in per capita consumption by 16 percent on average the year of the event (Thomas et al., 2010). It was also the case in the Philippines. Anttila-Hughes and Hsiang (2013) showed that households

reduced their consumption after typhoons. According to Skoufias et al. (2019), typhoons in

the Philippines led to negative effects on consumption levels, and in particular food. After

Typhoon Milenyo, many households decreased their consumption in the short run, of rice for

example, which is the primary food in the Philippines (Sakai et al., 2017). For Bangladesh,

Del Ninno et al. (2003) showed that poor households decreased their consumption after the

1998 floods in the short run. Paul and Routray (2010b) found that households decreased their

consumption after cyclones and induced surges in coastal Bangladesh. According to Mottaleb

et al. (2013), Cyclone Aila caused a significant decrease in paddy consumption of own-

produced paddy. Then, the Great Hanshin‐Awaji earthquake in Japan had a negative effect on

household consumption. If households encountered higher losses, they were more likely to

reduce their consumption compared to households that faced fewer losses (Sawada and

Shimizutani, 2008). Van den Berg and Burger (2008) found a drastic decline in consumption

after Hurricane Mitch in Nicaragua. The majority of households counted on internal strategies

to deal with the consequences, by decreasing consumption after the losses in assets (main

strategy) and not selling assets. Then, the earthquakes in El Salvador caused a decline in

current and future consumption because of the fall in income and assets (Baez and Santos,

2008). Ghorpade (2012) studied poor households and how they react to income shocks after

natural disasters. Some households reduced their consumption on non-necessary items,

investments, schooling of children, healthcare and food. However, Helgeson et al. (2013)

found that decreasing consumption was not used frequently in rural Uganda.

6.1.4 Assets

We can now see the importance of assets. Assets make households more resilient to shocks according to Briguglio et al. (2009), Bruneau et al. (2003) and Davies et al. (2013). Liquid assets might help rural households after the changes in consumption (Wainwright and Newman, 2011).

Some households sold their assets in the aftermath of natural disasters. In the short run, some households sold their productive assets as a coping mechanism, in order to sustain a minimum level of consumption (Rosenzweig and Binswanger, 1993; Rosenzweig and Wolpin, 1993;

Fafchamps et al., 1998; Kazianga and Udry, 2006). It was also the case in Bangladesh: in the disaster-prone-areas (Shahabuddin and Ali, 2006), after the 1998 floods (Del Ninno et al., 2003), rural poor households sold portable assets after the 1998 floods (Shah, 1999), after the floods in two villages (Bannabari and Suvagacha) (Paul and Routray, 2010a) and after

cyclones and induced surges in coastal Bangladesh (Paul and Routray, 2010b). Abdullah et al.

(2016) showed that 15 percent of households sold their assets after Cyclone Aila. Also,

households in El Salvador liquidated critical productive assets after the earthquakes (Baez and Santos, 2008). Helgeson et al. (2013) found that households in rural Uganda sold their assets, such as lands and livestock (selling livestock was the most frequently coping strategy used).

Finally, Ghorpade (2012) discovered that poor households accumulated and sold assets, such as livestock, in order to deal with shocks.

6.1.5 Reaction in the labour market

Households react after changes in the labour market.

Some people tried to be employed in other areas than agriculture. After the 1998 floods in Bangladesh, some jobs were created to repair the damaged houses and in fishing and some poor households moved to urban areas to find jobs in the non-farm sector, however the extent of migration was not significant (Shah, 1999). It was also the case after Typhoon Milenyo in the Philippines: a coping strategy after the event was working in non-farm employment (Sakai et al., 2017) and rural households used non-farm employment as a coping strategy after

Milenyo because non-farm income was not affected (Sawada et al., 2009). After natural

disasters in Mexico, farmers had sources of off-farm income in the short run (Saldaña- Zorrilla, 2008).

Some households can decide to work more. Van den Berg and Burger (2008) found that, after Hurricane Mitch (Nicaragua), a coping strategy was to work longer hours. Carter et al. (2007) discovered that households could increase their work time and reduce leisure, in order to maintain their consumption levels, after natural disasters in Ethiopia and Honduras.

6.1.6 Communities

Some households benefited from help within the communities. Local support networks was a coping strategy that households used after natural disasters in Vietnam (Arouri et al., 2015).

According to Paul and Routray (2010b), households benefited from help within the

community, after cyclones and induced surges in coastal Bangladesh. Moreno (2018) showed that communities were active agents after the earthquake and tsunami in Chile and they could cope and recover from natural disasters thanks to resilience capacities, such as social

networks, organisation, solidarity, cooperation, trust, local knowledge and participation. After Hurricane Katrina in the United States, communities used different capacities, just like local knowledge, resources, strategies to survive and recover. Ghorpade (2012) explained that solidarity within the communities is important in the aftermath of natural disasters for poor households, with for instance gifts, transfers, cash remittances and advances of money.

6.1.7 Other examples of coping mechanisms

There exist other examples of coping strategies that households used after natural disasters, for instance change in child schooling, change in fertility, change in labour supply in Vietnam (Arouri et al., 2015). Ghorpade (2012) found that some poor households increased their labour supply in reaction to income shocks after natural disasters. Saldaña-Zorrilla (2008) found that farmers adapted their production and made their infrastructure better (in the long run) after natural disasters in Mexico. According to Adeagbo et al. (2016), Nigerian

households adopted coping strategies, such as having a low standard of living.

6.2 Coping mechanisms of governments and institutions

Coping strategies of households are not enough, public interventions are needed to minimize the consequences of natural disasters (Ninno et al., 2013).

Governments and institutions can help households. In rural Vietnam, the government and other organizations (e.g., NGOs, United Nations) implemented programs and policies to attenuate the negative effect of natural disasters (Arouri et al., 2015). Also, the Vietnamese government and the World Bank built disaster mitigation infrastructure, through the

implementation of the “Community-based disaster risk management program” in the Mekong Delta (Trinh et al., 2021). For the Philippines, 49 percent of the households received a

government program in Pasay city (Israel and Briones, 2014). After Typhoon Milenyo, households received aid from local government and NGOs (Sawada et al., 2009; Sakai et al., 2017). After the Wenchuan earthquake, the Chinese government provided necessities and aid to reconstruct houses (Feng et al., 2016). Then, households received support from government and agencies, in the disaster-prone-areas in Bangladesh (Shahabuddin and Ali, 2006).

Ishtiaque and Nazem (2016) discovered that only 12 percent of households got food and clothing immediately after the disasters in Dhaka city and 8 percent of people received cash for the rebuilding of their houses. And the rest of the households did not receive any aid from the government or non-governmental organisations. According to Mueller and Quisumbing (2011), the government of Bangladesh implemented two programs after the 1998 flood: the Gratuitous Relief (GR) program (households in the flood-affected areas benefited from food aid) and the Vulnerable Group Feeding (VGF) program (for poor households in the flood- affected areas). Also, after the 1998 floods in Bangladesh, the government made “food, drinking water, medicines, shelter, and jobs available to the poor” (Shah, 1999). Alam et al.

(2018) found that the government, educational institutions and NGOs implemented coping strategies in disaster prone areas after tropical cyclones in the Kuakata coastal belt of

Patuakhali Bangladesh, such as construction of cyclone shelters, dams and sluice gates. They also raised awareness on disaster plans, such as first aid training programs. And Paul (2005) found that households received disaster relief after the tornado in the north of Bangladesh.

Then, the Indian government gave money to households in Mumbai and Chennai after rainfall

events, in order to deal with immediate needs, such as food and clothing. But, it only

represented about 10 percent of the overall losses that people experienced (Patankar, 2019).

For Mexico, the government aid was one of the most important financing sources (Saldaña- Zorrilla, 2008). Also, the Mexican government provided some relief for poor households after natural disasters in vulnerable regions (Saldaña-Zorrilla and Sandberg, 2009). After Hurricane Mitch, Nicaraguan and international NGOs gave support in the affected areas, such as food and psychological assistance. Multilateral organisations also helped households (Van den Berg and Burger, 2008). Martínez et al. (2020) found that the negative impacts on the Chilean labour market in the short run could be attenuated after a few years (in the long run) thanks to the implementation of recovery policies and programs. After L’Aquila’s earthquake, the Italian government provided food, temporary shelter for households (Ambrosetti and Petrillo, 2016). According to Adeagbo et al. (2016), all categories of Nigerian households had access to a government relief, with the aim to recover speedily.

Governments can help workers in particular. Venn (2012) studied natural disasters in OECD countries and found that following natural disasters, solutions are implemented for workers.

Governments can help them in order to preserve them and help them recover after natural disasters. It can be in the short run and in the long run. Workers can have income support, for instance unemployment benefits (there were more unemployment benefits after disasters in most countries), subsidies, creation of emergency jobs (e.g., about 20 000 emergency jobs were created in Chile), social work programmes, unemployment insurance and assistance, etc.

These solutions face challenges, for instance communication can be hard after natural disasters because people have been displaced and because of the damage of communication facilities. It is needed to have the internet or phones to know what benefits workers can have and to ask for those benefits.

7 Conclusion

The purpose of this essay was to survey the literature and find the economic effects of natural

disasters for households. Then, we looked deeper at the worst disasters, who is the worst

affected, where natural disasters happen the most and the coping mechanisms of households

and countries, in order to try to help us learn from environmental disasters. This study is

important because it summarizes the previous literature, and it allows us to learn more from previous disasters in order to prepare better for the future.

The results showed that household income declined in most studies, the income reduction can even persist in the long run. In general, households lost assets. Environmental disasters

usually increase poverty. Household welfare often decreased after natural disasters. Moreover, natural disasters generally have negative impacts on the labour market: on employment, unemployment, some workers lost days of work or even lost their jobs. Then, migration often increased after environmental disasters. Finally, disasters are expected to increase inequality.

As for the worst disaster, results differ: floods, cyclones, earthquakes, etc. Some households seem to be more affected, such as poor households and women. Asian countries experience the worst economic losses, and the Philippines, China, Japan and Bangladesh seem to be the countries where there are the most natural disasters.

Households implemented coping mechanisms, in order to recover from natural disasters.

Indeed, they have access to different post-disaster sources, such as help from relatives, friends and neighbours, public and private transfers, borrowing and credits. They also use their savings. And they can benefit from insurance. Besides, households reduced their expenditures and changed their consumption, for instance a reduction in consumption. Some households sold their assets (productive assets, portable assets, etc.). Moreover, many households tried to be employed in other areas than agriculture and some households decided to work more. Also, some households benefited from help within the communities. Finally, households use many other coping strategies. As for policy implications, we have seen that some governments helped some households, however not everyone has access to aid from their government.

Some households do not receive help in some countries and when governments help people, they do not help every household affected by a natural disaster.

This study is limited in the sense that it surveys only natural disasters and their implications as they have been previously studied in the literature. Not all disasters, economic consequences or coping mechanisms have been studied. Therefore, there could be larger gaps in the

literature. However, many of the sources come to similar conclusions, so I am confident that

the larger picture is reliable.

Future research in the area is necessary, indeed there is not a lot of research upon the

consequences of natural disasters for households and it could be interesting to find empirical evidence about the consequences. And it is important to further investigate the coping mechanisms, in order to better deal with natural disasters.

8 References

21st Century Challenges. (2015). Who are most vulnerable to natural hazards? [online]

Available at: https://21stcenturychallenges.org/who-are-most-vulnerable-to-natural-hazards/

[Accessed 13 May 2021].

Abdullah, A.N.M., Zander, K.K., Myers, B., Stacey, N. and Garnett, S.T. (2016). A short- term decrease in household income inequality in the Sundarbans, Bangladesh, following Cyclone Aila. Natural Hazards, [online] 83(2), pp.1103–1123. Available at:

https://link.springer.com/article/10.1007/s11069-016-2358-1 [Accessed 23 May 2021].

Adeagbo, A., Daramola, A., Carim-Sanni, A., Akujobi, C. and Ukpong, C. (2016). Effects of natural disasters on social and economic well being: A study in Nigeria. International Journal of Disaster Risk Reduction, [online] 17, pp.1–12. Available at:

https://www.sciencedirect.com/science/article/pii/S2212420915301370?casa_token=_- vTv6Bz5bYAAAAA:36cpigQU-

R7U2_POsuvaJO3zJXhq2lUnlHgtJ0CP9mIguh0YQK6rGWIqid8HFZedOiWmELlcpQ#s015 0 [Accessed 26 Mar. 2021].

Alam, M.Z., Halsey, J., Haque, M.M., Talukdar, M., Moniruzzaman, M. and Crump, A.R.

(2018). Effect of Natural Disasters and Their Coping Strategies in the Kuakata Coastal Belt of Patuakhali Bangladesh. Computational Water, Energy, and Environmental Engineering, [online] 07(04), p.161. Available at: https://www.scirp.org/html/1-2570183_87960.htm#txtF3 [Accessed 15 Apr. 2021].

Ambrosetti, E. and Petrillo, E.R. (2016). Environmental disasters, migration and

displacement. Insights and developments from L’Aquila’s case. Environmental Science &

Policy, [online] 56, pp.80–88. Available at:

https://www.sciencedirect.com/science/article/pii/S1462901115301039?casa_token=AZK1wp

TudQUAAAAA:uQxYC5TR53Dv2F1gjHmIB6i5po2KaJAKnzLPDVBvGlDClWj3MYUpZI ZYRMn9boktVm_B-bZog [Accessed 26 Mar. 2021].

Anttila-Hughes, J. and Hsiang, S. (2013). Destruction, Disinvestment, and Death: Economic and Human Losses Following Environmental Disaster. [online] SSRN. Available at:

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2220501 [Accessed 26 Mar. 2021].

Arouri, M., Nguyen, C. and Youssef, A.B. (2015). Natural Disasters, Household Welfare, and Resilience: Evidence from Rural Vietnam. World Development, [online] 70, pp.59–77.

Available at:

https://www.sciencedirect.com/science/article/pii/S0305750X1400415X?casa_token=JgGLm NFBM_UAAAAA:LGC_y3PVgpV36UL5doapv7lAozmrWzREcTyKgZRR9QlKnzWJlITT8 1Z4CtFtJ3BaJT-R41clA [Accessed 26 Mar. 2021].

Baez, J.E. and Santos, I.V. (2008). On Shaky Ground: The Effects of Earthquakes on Household Income and Poverty. [online] Available at:

https://www.researchgate.net/profile/Indhira_Santos/publication/237312732_United_Nations _Development_Programme_Regional_Bureau_for_Latin_America_and_the_Caribbean_On_

Shaky_Ground_The_Effects_of_Earthquakes_on_Household_Income_and_Poverty/links/004 6353baa1085cbbd000000/United-Nations-Development-Programme-Regional-Bureau-for- Latin-America-and-the-Caribbean-On-Shaky-Ground-The-Effects-of-Earthquakes-on- Household-Income-and-Poverty.pdf [Accessed 27 Apr. 2021].

Bui, A.T., Dungey, M., Nguyen, C.V. and Pham, T.P. (2014). The impact of natural disasters on household income, expenditure, poverty and inequality: evidence from Vietnam. Applied Economics, [online] 46(15), pp.1751–1766. Available at:

https://www.tandfonline.com/doi/full/10.1080/00036846.2014.884706?casa_token=rLrLw6O sDIQAAAAA%3AmFZe84FomW4DMV3xaCVmNjbJtXXPYzvhSwILkBgc9HdgQpVr0kjP JOUze-MPArrJjN8ThiXEc [Accessed 26 Mar. 2021].

Carter, M.R., Little, P.D., Mogues, T. and Negatu, W. (2007). Poverty Traps and Natural Disasters in Ethiopia and Honduras. World Development, [online] 35(5), pp.835–856.

Available at: https://www.sciencedirect.com/science/article/pii/S0305750X07000149

[Accessed 26 Mar. 2021].