Policy Research Working Paper 8723 Measuring Natural Risks in the Philippines Socioeconomic Resilience and Wellbeing Losses Brian Walsh Stephane Hallegatte Climate Change Group & Social, Urban, Rural and Resilience Global Practice January 2019 Policy Research Working Paper 8723 Abstract Traditional risk assessments use asset losses as the main Philippines is estimated at US$3.9 billion per year, more metric to measure the severity of a disaster. This paper than double the asset losses of US$1.4 billion. Second, proposes an expanded risk assessment based on a frame- the regions identified as priorities for risk-management work that adds socioeconomic resilience and uses wellbeing interventions differ depending on which risk metric is losses as its main measure of disaster severity. Using a new, used. Cost-benefit analyses based on asset losses direct risk agent-based model that represents explicitly the recovery reduction investments toward the richest regions and areas. and reconstruction process at the household level, this A focus on poverty or wellbeing rebalances the analysis risk assessment provides new insights into disaster risks in and generates a different set of regional priorities. Finally, the Philippines. First, there is a close link between natural measuring disaster impacts through poverty and wellbe- disasters and poverty. On average, the estimates suggest ing impacts allows the quantification of the benefits from that almost half a million Filipinos per year face transient interventions like rapid post-disaster support and adaptive consumption poverty due to natural disasters. Nationally, social protection. Although these measures do not reduce the bottom income quintile suffers only 9 percent of the asset losses, they efficiently reduce their consequences for total asset losses, but 31 percent of the total wellbeing losses. wellbeing by making the population more resilient. The average annual wellbeing losses due to disasters in the This paper is a product of the Global Facility for Disaster Reduction and Recovery, Climate Change Group with the Social, Urban, Rural and Resilience Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The authors may be contacted at bwalsh1@worldbank.org and shallegatte@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team : & Keywords: natural risks, resilience, risk assessment, welfare, Philippines JEL: D15, D30, D63, D78, Q54, R11 Global Facility for Disaster Reduction and Recovery (GFDRR), World Bank Group 2 The Philippines is by some measures among the most disaster-affected countries in the world. The 100 million residents of the Philippines, along with their homes and livelihoods, are exposed to a wide variety of disasters, including typhoons, earth- quakes, floods, storm surges, and tsunamis. When major storms and earthquakes affect densely populated urban areas in the country, asset losses are regularly valued in the billions, with additional untold human costs. In addition, small events occur frequently across the 2,000 inhabited islands in this middle-income country. On November 8, 2013, Super Typhoon Yolanda (internationally referred to as Ty- phoon Haiyan) made multiple landfalls in the Eastern, Central, and Western Visayas regions of the Philippines, claiming nearly 6,300 lives and directly affecting more than 16 million individuals across nine regions. At least 1.1 million homes were dam- =95 aged or destroyed, and the government estimated total losses at US$2.2 billion (P billion), making Yolanda the most costly hurricane to affect the Philippines to date. Since Yolanda, various regions of the Philippines have been affected by numerous wind, earthquake, and flood events, both large and small. As these statistics make clear, the consequences of disasters in the Philippines ex- tend beyond the replacement costs of destroyed assets — or asset losses. Disasters have well-documented consequences for social and economic development agendas and outcomes, including especially inequality, agriculture, education, and health [1, 2, 3, 4, 5, 6, 7, 8]. Despite this growing body of research, risk assessments still typically adopt asset losses as a singular metric of disaster impacts. This is unfortunate, because asset losses obscure the relationship between disaster risk and poverty. By definition, wealthy individuals have more assets to lose, and therefore their interests dominate in risk assessments that are limited to asset losses. At the same time, asset losses do not measure many dimensions of disaster impacts that accrue to the poor: while they by definition have very little to lose, they also lack the resources and instruments to smooth income shocks while maintaining their consumption, and to recover and rebuild their asset stock. Therefore, the poor are 3 more likely than the wealthy to forego consumption of food, health, or education in order to finance their recovery, and to take longer to recover. To correct this bias, the initial Unbreakable report introduced the concept of well- being losses. While proportional to traditional asset losses, wellbeing losses account for people’s socio-economic resilience, including (1) their ability to maintain their consumption for the duration of their recovery, (2) their ability to save or borrow to rebuild their asset stock, and (3) the decreasing returns in consumption–that is, the fact that poorer people are more affected by a $1 reduction in consumption than richer individuals (see Figure 1). The analysis presented in this paper builds upon and expands the approach pro- posed in the initial report. It uses a new agent-based model along with detailed natural-risk and household survey data to examine the consequences of natural disas- ters on households, as measured by asset losses and alternative metrics. This analysis provides a multi-metric assessment of disaster risks at the regional level, using: (1) traditional asset losses; (2) poverty-related measures such as poverty headcount; (3) wellbeing losses, which provide a balanced estimate of the impact on poor and rich households; and (4) socio-economic resilience, an indicator that measures the ability of the population to cope with and recover from asset losses. This broad perspective is intended to complement traditional, more spatially detailed risk assessments in the Philippines. The first conclusion of this analysis is the close link between natural disasters and poverty in the Philippines, and this connection goes both ways. First, natural disas- ters are a cause of poverty: on average, estimates suggest that almost half a million Filipinos per year face transient consumption poverty due to natural disasters. And in several regions throughout northwestern Luzon, the number of individuals pushed by disasters below subsistence level represents at least 20% of chronic subsistence incidence. In these places, reducing risk is likely an efficient way of reducing poverty. But poverty also magnifies the impact of natural hazards by making people more vulnerable and less resilient. The bottom income quintile suffers only 9% of the national asset losses, but 31% of the total wellbeing losses. On average, the poorest quintile suffers from wellbeing losses that are 1.5 times larger than average individual loss in the country. As a result of the disproportionate impact on poor people, the 4 average annual wellbeing losses due to disasters in the Philippines is estimated at US$3.9 billion per year (3.3% of household expenditures), more than double the asset losses of US$1.4 billion (1.3% of household annual expenditures). A second conclusion is that priority interventions — both in spatial terms (where to act?) and sectoral terms (how to act?) — are highly dependent on which metric for disaster severity is used. The most important interventions will focus around Manila if asset losses are used as the main measure of disaster impacts, while regions like Bicol become priorities in terms of poverty incidence and wellbeing losses. And the least resilient region — the one that would struggle the most if it was affected by a dis- aster — is the poorest, ARMM. Further, one needs to use regional averages with care: our results show that the poorest people in the richest regions are almost as vulnera- ble as the poorest people in the poorest regions. An important consequence of these findings is that the choice of the metric used in risk assessments is not a technical question, but a political choice with significant implications for which interventions are desirable. Finally, the third conclusion of this work is that new metrics of disaster impacts — including poverty headcount, poverty gap, and wellbeing losses — can be used to quantify the value of interventions currently outside the traditional risk-management toolbox. Asset-informed risk-management strategies primarily focus on protection infrastructure, such as dikes, and the position and condition of assets, for instance with land-use plans or building norms. Wellbeing-informed strategies can utilize a wider set of available measures, such as financial inclusion, private and public insurance, disaster-responsive social safety nets, macro-fiscal policies, and disaster preparedness and contingent planning. Even if they do not reduce asset losses, these measures can bolster communities’ socio-economic resilience, or their capacity to cope with and recover from asset losses when they occur, and reduce the wellbeing impact of natural disasters. Beyond these policy conclusions, this paper also describes for the first time an agent-based model developed to better understand the impacts of natural disasters on diverse households and their paths to recovery. Its primary innovation is the use of the Family Income & Expenditure Survey (FIES) to disaggregate expected asset losses among representative households, resulting in a measurement of asset losses, 5 Figure 1: Traditional risk assessments evaluate asset exposure and vulnerability to hazards to determine expected asset losses. The Unbreakable model additionally incorporates the socio-economic resilience of the communities to predict wellbeing losses. poverty impacts, and wellbeing losses by income quintile and region in the country (or many other possible categories, including gender, education level, and sector of employment).1 The analysis starts from hazard- and asset-class-specific exceedance curves at the provincial level from the Government of the Philippines Department of Finance Catas- trophe Risk Model (DFCRM)[9, 10]. The first significant innovation of this approach is to distribute these losses among the representative households of the FIES. This step is based on their asset vulnerability, estimated from available household charac- teristics (i.e., from housing construction materials and condition).2 The second main innovation of the model is to explicitly represent disaster recon- struction dynamics at the household level using an agent-based approach in which (1) each household acts rationally to minimize its wellbeing losses, and (2) households interact through firms’ activities and government budgets. The model specifies each household’s unique reconstruction and savings expenditure rate, assuming house- holds optimize the fraction of income they dedicate to repairing and replacing their assets, at the expense of immediate consumption. For instance, people close to the 1 In geographical terms, the current version of the FIES is representative only at the regional level. The 2018 FIES will make it possible to perform the same analysis at the provincial level. 2 Unfortunately, the FIES does not currently include the geolocations of households. 6 subsistence level cannot set aside much of their income to rebuild their assets with- out experience large wellbeing losses, and may therefore take longer to recover. In extreme cases, they may even be trapped in poverty, generating large wellbeing losses going well beyond the few years that follow a disaster [11, 12]. With this approach, we are able to develop detailed country risk profiles that in- corporate sub-regional variations in hazard, exposure, asset vulnerability, and socioe- conomic resilience and wellbeing losses, to be used as an input in a prioritization process at as fine a scale as data allow. We can also demonstrate the benefits of formal and informal risk sharing mechanisms, including post-disaster support and private remittances, in terms of increased resilience and reduced impacts of natural disasters on lives and livelihoods in the Philippines. The paper is organized as follows. Section 2 provides an overview of asset risk due to wind, precipitation floods, storm surge, and earthquake events in each of the 17 regions. In Section 3, we overlay this information with household-level in- come and asset vulnerability data from the 2015 Family Income and Expenditures Survey (FIES) to develop finely-grained estimates of income and consumption losses to disasters. Based on these results, we estimate the number of households in the Philippines facing transient income or consumption poverty each year due to natural disasters. In Sections 4 to 6, we quantify wellbeing risk at the national and regional levels, respectively. In Section 7, we examine specifically the risk to assets and risk to wellbeing of the poorest quintile in each region. Section 8 describes policy simula- tions, and Section 9 presents conclusions. A technical appendix provides details on the methodology and data, and the equations of the freely-available model. Typically, risk assessments incorporate information on the hazards (the natural occur- rence of destructive events such as severe winds, surges, floods, and earthquakes); exposure (the value of natural and built assets that might face a destructive event); and vulnerability (the expected consequences to exposed assets when a destructive 7 event occurs) of the targeted area. Together, these three dimensions describe average annual asset losses in the area of interest. Table 1 on the following page displays average annual asset losses, by hazard type, for each region in the Philippines.3 Already, we see that disasters in NCR (metropoli- tan Manila) and in CALABARZON (region IVA) cost over US$300 million per year, on average. In NCR, these losses are driven primarily by precipitation flooding, while wind events are responsible for over 50% of annual losses in CALABARZON. Significant differences across regions and disaster type are explained by variations in the hazard, exposure, and vulnerability of each part of the country. For example, while catastrophic typhoons can strike any part of the country, the northeastern coast of Luzon (regions II, IVA, and V) and the Eastern Visayas (region VIII) are the most frequently affected–that is, these regions face elevated typhoon hazard. Further, all the regions facing asset risk in excess of US$100 million are in mainland Luzon, the wealthiest and most developed part of the country (elevated exposure). By contrast, asset risk in the whole of Mindanao (regions IX, X, XI, XII, XIII, and ARMM) accounts for just 5% of total losses due to the relative infrequency of major events (low hazard) and high poverty in these regions (low exposure, high vulnerability). Figure 2 on page 9 maps multihazard asset risk in the Philippines. On the left, we represent losses in millions of dollars (as in the rightmost column in Table 1). On the right, the same results are presented as a percentage of each region’s AHI. Generally, the same regions (i.e., metropolitan Manila and surrounding parts of Luzon) are highlighted in both representations of asset risk. However, in the map on the right, Cagayan Valley and Bicol (regions II and V, respectively) have replaced NCR and CALABARZON as the most heavily-impacted regions. The simple comparison in Figure 2 illustrates an essential point: different metrics can lead to different disaster risk “hotspots" — and, therefore, priorities. At a mini- 3 Comprehensive asset risk is a direct output of the AIR catastrophe model [9, 13], but this analysis is particularly interested in the impacts of disasters on household assets, consumption, and welfare. One issue we face is the difference between the Gross Domestic Income (GDI) derived from national accounts and the aggregated household income (AHI) calculated from household surveys (in this case, FIES). As is well known, the latter tend to report lower incomes than national accounts [14]. In this analysis, we work based on the AHI. Therefore, the expected losses in Table 1 have been scaled by AHI as a fraction of the nominal regional productivity (GRDP), a factor of 0.43 on average (cf. Tab. 6). 8 Asset losses [mUS$ per year] Region EQ HU PF SS All hazards NCR 68.6 67.1 169.7 1.4 306.8 IVA - CALABARZON 71.5 156.3 67.2 9.8 304.9 III - Central Luzon 56.2 119.2 54.8 3.4 233.5 V - Bicol 11.2 68.4 7.8 27.0 114.4 I - Ilocos 37.7 62.4 13.1 0.8 114.0 II - Cagayan Valley 9.7 88.7 9.5 6.0 113.8 VIII - Eastern Visayas 11.3 34.1 2.1 13.2 60.7 VII - Central Visayas 8.9 15.0 25.8 9.9 59.6 VI - Western Visayas 19.3 11.1 8.3 7.3 46.0 IVB - MIMAROPA 6.1 11.1 1.7 0.8 19.8 CAR 3.5 9.0 6.2 0.0 18.7 XI - Davao 12.6 0.6 0.9 0.0 14.1 XIII - Caraga 9.9 3.0 0.5 0.5 13.9 XII - SOCCSKSARGEN 7.2 0.1 0.1 0.0 7.3 X - Northern Mindanao 3.3 2 .4 1.3 0.2 7.1 ARMM 4.5 0 .3 0.3 0.0 5.1 IX - Zamboanga Peninsula 3.2 0 .2 0.2 0.0 3.6 National total 344.7 648.9 369.4 80.4 1,443.3 Table 1: Expected annual asset losses in millions of US$ from earthquakes (EQ), hurricanes (HU), precipitation floods (PF), and storm surges (SS) for each region. mum, this suggests that asset risk does not give a complete picture of disaster impacts in the Philippines. Both maps in Figure 2 describe mostly what happens to the wealth- iest regions — and wealthiest people in these regions — because they are based on aggregate loss data, and wealthy regions and people have the most to lose. Although asset losses in ARMM total just US$5 million per year, disaster responses (and, there- fore, disaster risk strategies) should in some way account for the 54% poverty rate in ARMM (versus 4% in NCR). Clearly, a more spatially disaggregated approach will be required. But moreover, we need to understand and develop metrics for those dimensions of disaster impacts that accrue to the poor. To these ends, the rest of this analysis goes deeper in the distributional analysis, moving from the regional scale to the household level. In the next section, we merge regional asset loss data with household-level socioeconomic characteristics to examine disaster impacts on household income and consumption. This novel approach will allow us to develop estimates of the number of individuals pushed into transient poverty each year by disasters. , , 9 Figure 2: Total risk to assets (expected annual losses) from earthquakes, hurricanes, precipi- tation floods, and storm surges for each of the 17 regions of the Philippines. At left, annual expected asset losses are expressed in US$. At right, losses are shown as a percentage of regional AHI. , , The technical details of asset loss disaggregation to the household level are discussed in teh technical appendix to this report. In this section, we present the main insights generated by the union of the DFCRM with the FIES, focusing on how individual households’ socioeconomic characteristics can mitigate or magnify the impact of dis- asters. , , 10 Income losses When disasters damage or destroy the assets on which individuals rely for their liveli- hood — including not only their own shop or field, but also somebody else’s factory — affected households face income losses. Some may receive extraordinary public as- sistance or additional remittances, which supplement their regular income while they rebuild. Insofar as they incorporate some of the socioeconomic characteristics that in- fluence households’ recovery pathways and describe better than asset losses the real impact on wellbeing, net income losses are a useful metric of disaster impacts. To provide a historical example: when Super Typhoon Yolanda made landfall in the Eastern Visayas (region VIII), it caused an estimated US$1.4 billion in damages to that region alone. In terms of asset losses, then, Yolanda was a roughly 100-year typhoon (including damage from wind, storm surge, and precipitation flooding) in the region. The top panel of Figure 3 on the following page illustrates the expected impact of a Yolanda-like wind event on individual incomes in the Eastern Visayas region. The black outline indicates the regional income distribution as reported in FIES, while the red histogram illustrates the expected income distribution immediately following a Yolanda-like event in the region. This distribution shows a mode around US$350 per person, per year, with nearly 40% of the population living below the poverty line. The large bar on the right shows the number of people with incomes higher than US$2,500 per year; the poverty and subsistence lines are indicated by the dotted lines around US$350 and US$450 per year, respectively. For this Yolanda-like event, the wind destruction alone is expected to push over 160,000 individuals into income poverty in Eastern Visayas, and over 170,000 below the subsistence income level (4% of the regional population). , , 11 500 100-year hurricane in VIII - Eastern Visayas Subsistence line Increase of 172,700 (3.8% of regional pop.) in income subsistence 400 Poverty line Increase of 160,300 (3.5% of regional pop.) in income poverty 300 Population (,000) Pre-disaster income (FIES data) Post-disaster income 200 (modeled) 100 0 0 500 1000 1500 2000 2500 Income [USD per person, per year] 500 Subsistence line Increase of 231,300 (5.1% of regional pop.) in consumption subsistence 400 Poverty line Increase of 176,800 (3.9% of regional pop.) in consumption poverty 300 Population (,000) Pre-disaster consumption (FIES data) Post-disaster consumption 200 (modeled) 100 0 0 500 1000 1500 2000 2500 Consumption [USD per person, per year] Figure 3: Expected impact of Yolanda-like (100-year) hurricane on per capita income (top) and consumption (bottom) in the Eastern Visayas (region VIII). Income losses take into account the lost productivity of disaster-affected assets, while consumption losses additionally include reconstruction costs, post-disaster support, and savings. , , 12 Consumption losses Income losses provide new insight into disaster impacts and are closer than asset losses to the actual impact on people. However, they still do not take into account a range of characteristics and coping mechanisms that can mitigate or exacerbate the effects of disasters on individual households. For example, we have not yet ac- counted for the reconstruction costs that directly-affected (and even those indirectly affected) households must pay to rebuild their assets after a disaster. Further, many households have some amount of savings to be used in case of a disaster, and wealth- ier households may benefit from formal and informal post-disaster transfers [15, 16]. These costs and resources all impact households’ consumption losses, or their (in)ability to maintain consumption when their income drops. The red histogram in the bottom panel of Figure 3 represents the household con- sumption distribution in the Eastern Visayas region immediately after a Yolanda-like hurricane in the region. After accounting for reconstruction costs and precautionary savings, the event is expected to generate a net increase of 177,000 individuals with consumption at or below the poverty line (4% of the regional population), and over 230,000 with consumption at or below the subsistence line (cf. Fig. 3). At the other end of the distribution, we note that there is no longer a discernible difference be- tween pre- and post-disaster consumption for households whose pre-disaster income is at least US$2,500 (compare to income losses). This suggests that the poor struggle to cope with lost income, while the wealthy are able to use savings and other instru- ments to maintain their consumption, even when they are affected by large disasters. Multihazard risk and chronic poverty On average, inclusive of all hazards and regions, we estimate that almost half a mil- lion Filipinos per year face transient consumption poverty due to natural disasters. This is equivalent to 2.2% of national poverty incidence, though the sub-national re- sults indicate significant regional variation (cf. Table 2 on the next page). For example, in several regions throughout northwestern Luzon, including NCR, Cagayan Valley, , , 13 CALABARZON, and Central Luzon, the number of individuals pushed into subsis- tence represents at least 20% of subsistence incidence, as measured by FIES. Although we do not model economic growth or other pathways out of poverty, these results in- dicate that natural disasters are major drivers of extreme poverty in certain regions of the Philippines. In these areas, disaster risk management strategies can be a highly efficient tool to decrease poverty and subsistence incidence, either by reducing asset losses or by enhancing households’ ability recover from disasters. Consumption poverty impacts of natural disasters Poverty Subsistence Increase % of regional Increase % of regional Region [thousands] incidence [thousands] incidence IVA - CALABARZON 107 8.3 76 23.1 III - Central Luzon 97 7.8 69 20.6 NCR 82 16.6 36 43.6 V - Bicol 55 2.5 65 9 .2 II - Cagayan Valley 26 4.7 23 25.7 I - Ilocos 26 3.8 16 9 .4 VII - Central Visayas 21 1.0 24 2.8 VIII - Eastern Visayas 17 1.0 22 2.9 VI - Western Visayas 15 0.9 13 2.3 XII - SOCCSKSARGEN 6 0.4 4 0.4 XIII - Caraga 5 0.4 6 1.4 IVB - MIMAROPA 4 0.6 4 1 .5 CAR 4 1.1 4 3 .4 XI - Davao 4 0.4 4 1 .1 X - Northern Mindanao 2 0.1 2 0 .2 ARMM 2 0.1 3 0 .4 IX - Zamboanga Peninsula 1 0.1 1 0 .2 Total 473 2.2 375 4 .6 Table 2: Annual, regional impacts of natural disasters on consumption poverty and subsis- tence incidence. Results are expressed in thousands of individuals, and as percent- ages of the FIES 2015 regional poverty and subsistence rates. Shifts are defined as the number of individuals in consumption poverty or subsistence immediately after a disaster occurs, less the pre-disaster headcount. Poverty incidence is inclusive of subsistence. Figure 4 on the following page maps the annual consumption poverty impacts of multihazard exposure in each region. The map on the left indicates disaster-related poverty incidence as a percentage of the total population in each region, while the map on the right describes disaster-related poverty incidence relative to the FIES , , 14 Figure 4: These maps indicate natural disaster-related increases in the number of Filipinos expected to face poverty or subsistence for any duration each year. At left, map colors indicate new poverty incidence as a percentage of regional population. At right, map colors indicate new poverty incidence as a percentage of regional poverty rate (inclusive of subsistence). estimate of chronic poverty in each region. These relationships provide additional metrics by which to measure the human impacts of natural disasters, and can help policy makers to understand the link between natural disasters and poverty. , , 15 Socioeconomic characteristics and disaster recovery dynamics Asset losses (and the foregoing income poverty analysis) describe individual house- holds’ status in the instant after a hazard occurred, and provide useful insights into how governments and other first responders should target humanitarian relief. How- ever, another important question is whether disaster-affected households become mired in poverty or achieve a speedy recovery. In contrast to asset losses, income and consumption losses can inform on the temporal dimensions of disaster impacts. Returning to the Yolanda-like event, Figure 5 on the next page maps the expected time to recover 90% of the assets destroyed when a 100-year wind event strikes each region. Unsurprisingly, the map indicates that metropolitan Manila and surrounding parts of Luzon are the quickest to recover. In other words, although disasters are more frequent and costlier in these regions than elsewhere, the high density of house- holds and productive assets in northern Luzon may help disaster-affected individuals to recover on their own. When major disasters occur outside these areas, recovery is expected to proceed slowly if at all, and disaster strategies seeking to minimize re- covery times should focus on providing post-disaster support to these areas. Notably, this modeling result is largely independent of initial asset losses, expressing instead the overall socioeconomic resilience of each region. Of the estimated half million Filipinos pushed into transient consumption poverty by disasters each year, some 25,000 will still be in poverty 10 years later. In total, this complex picture suggest that the costs of natural disasters are far-reaching for certain households, communities, and regions. Accounting for loss heterogeneity and households’ different abilities to reconstruct suggests therefore the existence of a long-term impact of disasters on income, which may be difficult to detect in aggre- gate economic data because of the small income of the people suffering from these long-term effects, but have been documented after some large-scale disasters in other countries [12, 11, 17, 18]. This heterogeneity could help explain why some studies have found long-term impacts of disasters on growth [19, 20, 21], while others have not [22, 23, 24, 25]. As discussed, the asset loss map in Figure 2 highlighted the impacts of disasters on the richest people. In contrast, metrics that focus on poverty and reconstruction (e.g., , , 16 Figure 5: Map of time to recover 90% of assets destroyed in 100-year hurricane. Figure 4 on page 14 and Figure 5) reveal how poor Filipinos experience and cope with shocks, and are therefore essential inputs to anti-poverty development policies. Still, poverty headcounts do not provide insight into how households already in poverty — nor those whose income does not drop below the poverty line — are impacted by disasters. This is an important limitation, as the foregoing discussion is meant to be as inclusive and integrative as possible, without discounting disaster-related costs to the wellbeing of any households. This is why we now introduce the concept of 17 wellbeing losses, which can be used to measure disaster impacts on all households, without creating a bias toward the richer ones. The foregoing has shown that, even if disaster-affected households suffer identical asset losses, their consumption losses and recovery time will vary according to their socioeconomic status and the resources available to them (for instance because insur- ance, savings, remittances, and public support help some households to smooth the consumption shocks). Going further, $1 in consumption losses can have very different consequences for individual households, depending on their income. In particular, while rich households will be able to spend down their savings or cut on luxury consumption, poorer households will often have to cut on basic needs and essential consumption like food, health, or education, threatening their health, human capital, and long-term prospects. Disaster risk strategies and budgets should account for these differences, and well- being losses do precisely this. While $1 in asset or consumption losses affects a poor individual more than a rich one, wellbeing losses are defined such that a $1 wellbeing loss affects a rich and a poor individual equally. Wellbeing losses are calculated from consumption losses using a classical “welfare function.” This operation translates into wellbeing the value of a household’s consumption at each point in its unique re- covery, with decreasing returns to represent the fact that increasing consumption by $1 increases more the wellbeing of a poor individual (compared with a rich person).4 The difference in the wellbeing generated by $1 of consumption is a simple proxy for the continuum from survival consumption (the very first units of consumption that have the largest impact on wellbeing) to luxury consumption (which enhances wellbeing less and less). 4 Here, the unit of analysis is the households. It would be useful to do it at the individual level, to uncover intra-household distributional effects, linked to gender or age. In the absence of within- household data, we make the strong assumption that pre-disaster consumption and disaster losses are distributed equally per capita within each household. 18 Wellbeing losses integrate each household’s consumption losses over the duration of its recovery, and give more weight to the consumption losses experienced by poor people than to the losses experienced by richer people. In this way, wellbeing losses account for socioeconomic differences among households, and correct for the pro- wealthy bias inherent in asset losses without relying on binary thresholds like the poverty line. As a metric, they capture more fully the costs of disasters and the benefits of prospective DRM investments than do asset losses. Therefore, wellbeing- informed strategies are not merely more equitable, but more cost-effective than asset- informed strategies. As discussed in Section 2 on page 6, we estimate annual asset losses to households from natural hazards (inclusive of hurricanes, precipitation floods, storm surge, and =72 billion, cf. Table 3 on earthquakes) in the Philippines at over US$1.4 billion (P page 21), equivalent to 1.3% of household annual expenditures. Wellbeing losses are =193 billion; 3.3% of household expenditures) per much higher, at US$3.9 billion (P year. In other words, the real impact of disasters in the Philippines is equivalent to a decrease in national consumption by almost US$4 billion.5 The analysis has moved from asset to income, consumption, and wellbeing losses, in- corporating additional relevant household-level socioeconomic characteristics at each stage. To summarize these developments, we return to the traditional risk-assessment framework (cf. 1 on page 5). The three traditional components (i.e., hazard, exposure, and vulnerability) predict asset losses, but not the capacity of affected populations to cope with and recover from these losses. In response to this theoretical and practical shortcoming, we have effectively added a fourth component, called socioeconomic re- silience, which we now define as the ratio of expected asset losses to wellbeing losses. 5 More precisely, the impact on wellbeing is equivalent to a = P193 billion decrease in consumption that would be optimally shared across households in the country and across time. Or it is equivalent to a = P193 billion decrease in consumption that would be equally shared across households in the country, if all households had the same income (i.e. if all inequality had disappeared). 19 At the national level, expected asset and wellbeing losses total US$1.4 billion and US$3.9 billion, respectively. Therefore, the wellbeing impact of disaster losses in the 3 .9 Philippines is 170% ( 1 .4 = 2.7) higher than asset losses. On average, every $1 in asset losses is equivalent to a $2.70 consumption loss, as experienced by a household earn- ing the national average income. Alternatively, we can say that the socioeconomic resilience of the Philippines is 37%.6 Although this is an informative result, the na- tional average of socioeconomic resilience belies significant regional variation in the capacity of Filipino households to cope with and recover from disasters. Figure 6 on the following page maps socioeconomic resilience to disasters at the regional level. Among all regions, metropolitan Manila enjoys the highest resilience (61%). This is due not just to its overall wealth and poverty rate of just 4%, but also to its high degree of financial inclusion and social protection coverage (30% higher than the national average). On the other hand, despite these advantages, wellbeing losses to disasters in metropolitan Manila are still 64% ( 0.1 61 = 1.64) higher than asset losses, and most of these losses still accrue to the poorest residents of the capital region. On the other end of the resilience ranking, ARMM has the lowest resilience among all regions (15%). Low annual asset losses in ARMM indicate that major events are rare (low hazard) and that the total value of assets in the region is relatively small (low exposure), but resilience is low, and additional policies and measures should be implemented to help this fragile region cope when disasters inevitably occur. In addition to these diagnostics, the extension of the traditional hazard-exposure- vulnerability framework to include socioeconomic resilience has the benefit of ex- panding the disaster risk management "toolbox." Traditional risk-management strate- gies seek to mitigate hazards as well as the exposure and vulnerability of assets. Other tools (e.g., post-disaster support, financial inclusion, private and public insur- ance, and sovereign contingent credit) are not easily included in risk assessments 6 This estimate is significantly lower than that from the Unbreakable report, and cannot be directly com- pared, considering the difference between the models. The difference is explained by the improved consideration of distributional impacts (using the full survey instead of only two categories of house- holds) and the explicit representation of the reconstruction pathway. The difference confirms the need to model the reconstruction pathway in a dynamic manner and to include the impact of short-term consumption drops. 20 Figure 6: Regional map of multihazard socioeconomic resilience to disasters. because they do not affect asset losses. In a wellbeing-informed framework, however, the benefits of these interventions become obvious and quantifiable: to the degree that they allow households to maintain a healthy degree of consumption while they rebuild, these interventions increase socioeconomic resilience and reduce wellbeing losses to disasters. 21 The annual risk to assets, risk to wellbeing, and socioeconomic resilience for each region of the Philippines are summarized in Table 3. In absolute terms, risk to house- hold assets is highest in metropolitan Manila (NCR) and in CALABARZON. Asset =30 billion) per year, over 40% of the losses in these two regions total over US$610M (P expected losses in the entire country. The asset and population density in these areas make risk management interventions effective, but also expensive. Socio-economic Asset losses resilience Wellbeing losses Region [mUS$/year] [%] [mUS$/year] IVA - CALABARZON 304.9 45.8 665.7 V - Bicol 114.4 18.9 605.3 III - Central Luzon 233.5 42.4 551.1 NCR 306.8 60.7 505.6 I - Ilocos 114.0 35.9 317.7 II - Cagayan Valley 113.9 39.3 289.7 VIII - Eastern Visayas 60.7 24.4 249.3 VII - Central Visayas 59.6 28.9 206.2 VI - Western Visayas 46.0 28.1 163.9 IVB - MIMAROPA 19.8 30.4 65.0 XIII - Caraga 13.9 22.8 60.9 XI - Davao 14.1 29.9 47.1 CAR 18.7 47.8 39.1 XII - SOCCSKSARGEN 7 .3 19.3 38.0 ARMM 5 .1 15.2 33.2 X - Northern Mindanao 7 .1 26.4 27.0 IX - Zamboanga Peninsula 3 .6 24.9 14.4 Total 1,443.3 37.2 3,879.2 Table 3: Asset losses, socio-economic resilience, and wellbeing losses for the entire population of each region of the Philippines. Asset and wellbeing losses are denominated in USD, and socio-economic resilience is the ratio of asset losses to wellbeing losses in each region. When disaster impacts are measured in wellbeing losses, an alternative set of re- gional priorities emerges. In particular, Table 3 indicates that Bicol and CALABAR- ZON suffer the greatest losses each year. In absolute terms, CALABARZON suffers =31 billion) per year, equiv- the highest wellbeing losses, surpassing US$665 million (P alent to 3.4% of AHI in the region. These losses are due to elevated exposure to 22 hurricanes and storm surges (51% and 22% of annual regional asset losses, respec- tively), which result in major wellbeing losses despite an above average value for socioeconomic resilience (45%) and a low regional poverty rate (just 9%). In Bicol, where asset losses are much lower, wellbeing losses are driven by a very high chronic poverty rate (36%), low household savings (60% of the national average, across all deciles), and low social transfer receipts (75% of national average). Overall, wellbeing losses are less geographically concentrated than asset losses, and this result is indicative of the magnitude of the challenge facing disaster managers in the Philippines. At the same time, it signifies new opportunities to mitigate disaster risk and chronic poverty by investing in socioeconomic resilience outside the most developed parts of Luzon. Because our results are based on information at the household level, we are able to assess asset and wellbeing losses not just at the regional level, but also for different income groups.7 Table 4 on page 24 lists annual asset and wellbeing losses for the poorest quintile in each region. Across all regions, the asset losses of the poorest =6.3 billion per year), or just 9% of total asset losses. On 20% are US$125 million (P the other hand, the wellbeing losses of the poorest quintile give a better sense of their experience of disasters: the wellbeing losses of the poorest quintile are valued =61 billion) per year, or 31% of total wellbeing losses. This means at US$1.2 billion (P that, on average, individuals in the poorest quintile suffer wellbeing losses that are 50% larger than the average individual loss in the country. Critically, Table 4 shows that the regional variations in socioeconomic resilience decreases significantly when we narrow our focus to the poorest Filipinos. While NCR has the highest overall resilience (61%), the poorest households in the capital have a resilience of only 17%. To wit: on average, the poorest residents of NCR lose 7 It is also possible to look at different subgroups in the country or at the regional scale (e.g., per occupation, head of household gender, household size, ethnic background or religion, social transfer enrollees), within the limits of the representativeness of the household survey. 23 Figure 7: Expected (average annual) wellbeing losses from earthquakes, hurricanes, precipita- tion floods, and storm surges for each region of the Philippines. Losses are shown in US$ (left) and as a fraction of regional aggregate household income (AHI, at right). US$9.50 per person, per year to natural disasters, but this equates to nearly US$60 per person and per year in wellbeing losses (cf. Table 5 on page 25). In terms of disaster impacts and recovery prospects, this result suggests that the poorest Manileños have more in common with the poor in other regions than with their wealthier neighbors. This is an important caveat to regional-scale assessments: the aggregate wealth of the capital region does not imply that its poorest residents are well-protected from or resilient to natural disasters. To the contrary, the low resilience of the poor to asset losses, combined with the large contribution of disasters to poverty incidence in 24 Socio-economic Asset losses Resilience Wellbeing losses Region [mUS$/year] [% of total] [%] [mUS$/year] [% total] IVA - CALABARZON 26.2 8.6 11.9 219.8 40.6 V - Bicol 13.1 11.4 6.4 204.5 38.1 III - Central Luzon 19.2 8.2 11.4 168.6 36.9 NCR 23.8 7.8 17.5 136.0 35.1 I - Ilocos 10.5 9.2 10.9 96.0 36.9 II - Cagayan Valley 10.5 9.2 12.1 87.2 38.2 VIII - Eastern Visayas 5.6 9.3 6.6 84.8 39.8 VII - Central Visayas 4.8 8.0 6.3 75.9 41.8 VI - Western Visayas 4.3 9.3 8.2 52.1 37.1 IVB - MIMAROPA 1.7 8.4 7.8 21.2 38.7 XIII - Caraga 1.2 8.9 6.0 20.6 38.9 XI - Davao 1.3 8.9 7.7 16.2 40.3 XII - SOCCSKSARGEN 0.6 7.7 4.1 13.9 39.9 CAR 1.1 6.0 10.5 10.7 34.1 X - Northern Mindanao 0.5 7.2 5.6 9 .2 38.2 ARMM 0.5 9.2 6.2 7 .5 24.6 IX - Zamboanga Peninsula 0.3 9.0 7.3 4 .4 35.4 Total 125.2 8.7 10.2 1228.5 38.1 Table 4: Natural disaster impact on the poor: risk to assets, socio-economic resilience, and risk to wellbeing for the poorest quintile (20%) in each region. Manila (cf. Tab. 2 on page 13), suggests that interventions to reduce asset losses and build resilience of the poor could be very effective for reducing disaster exposure and poverty incidence, particularly in Manila and other wealthy regions. The theoretical framework and model presented here are not limited to disaster risk diagnostics. They also allow for detailed analyses of the costs and benefits of diverse instruments and investments to reduce risks, or make the population better able to deal with them. While disaster response tools such as post-disaster cash transfers do not affect asset losses, their benefits can be measured by multiple poverty metrics (e.g., poverty headcount, poverty gap), reconstruction time for various households, and wellbeing losses. And since the model estimates losses at the household level, the 25 Per capita losses Q1 population Asset Well-being Region [thousands] [US$ per cap., per year] V - Bicol 1,206 10.86 169.51 II - Cagayan Valley 698 15.08 124.89 I - Ilocos 1,024 10.25 93.71 VIII - Eastern Visayas 906 6.21 93.55 IVA - CALABARZON 2,824 9.29 77.84 III - Central Luzon 2,218 8.65 75.98 Total 20,278 6.17 60.58 NCR 2,528 9.41 53.80 VII - Central Visayas 1,486 3.22 51.06 XIII - Caraga 541 2.29 38.04 IVB - MIMAROPA 614 2.70 34.56 VI - Western Visayas 1,539 2.77 33.84 CAR 355 3.15 29.98 XI - Davao 990 1.27 16.40 XII - SOCCSKSARGEN 912 0.62 15.17 ARMM 738 0.63 10.11 X - Northern Mindanao 940 0.54 9.75 IX - Zamboanga Peninsula 751 0.43 5.86 Table 5: Per capita risk to assets and wellbeing for the poorest quintile (20%) of each region. model also makes it possible to examine the distribution of these costs and benefits throughout the population. Post-disaster support in the Philippines In the Philippines, the Department of Social Welfare and Development (DSWD) is the lead agency for disaster response within the government’s National Disaster Risk Reduction and Management Plan (NDRRMP). In response to Yolanda, DSWD imple- mented a variety of SP and social welfare programs: distribution of in-kind relief items, cash transfers (unconditional and conditional), shelter, and community-driven development.8 Initially, the emphasis was on food and nonfood items (like mats, blankets, tarpaulins, hygiene kits, and clothing) to meet the immediate and urgent survival needs, plus temporary shelter assistance for displaced households. 8 A more complete description of the response to Yolanda is provided in the Shock Waves report [26]. 26 After immediate survival needs were addressed, DSWD delivered a number of cash-based response programs, such as Cash for Work, Cash for Building Liveli- hood Assets, and cash for shelter (Emergency Shelter Assistance)–then transformed into the Core Shelter Assistance Program to rebuild permanent housing. DSWD also temporarily removed all conditionality of the Pantawid Pamiliya Pilipino Pro- gram (4Ps), a usually conditional cash transfer program. In addition, at least 45 international humanitarian agencies implemented cash transfers (unconditional and conditional), partly delivered through the 4Ps infrastructure. Four agencies alone distributed around US$34 million, benefiting 1.4 million disaster-affected people. Modeling post-disaster support Here, we do not try to reproduce the very complex response to typhoon Yolanda, but instead we assess the benefit from an idealized post-disaster support (PDS) provided to the population after a disaster. As an illustrative exercise, we consider a simple post-disaster support system. In this system, all disaster-affected households receive a uniform cash payout, equal to 80% of the average asset losses suffered by the poorest quintile. This system ensures that poor people are compensated for a large fraction of their losses and assumes that all affected households are supported, while total costs remain acceptable. The cost of the program is distributed among all households in all regions via a flat tax on income. Returning to the Yolanda-like hurricane event: expected wind damage to house- hold assets in the Eastern Visayas region is valued at US$633 million. Wellbeing losses from the same event are valued at US$2,176 million, for a socioeconomic re- silience of 29%.9 In Figure 8 on page 28, we plot per capita asset and wellbeing losses, grouped by income quintile. The figure shows that the richest households lose the most assets, while the poorest households suffer the greatest wellbeing losses. 9 Here, we report expected asset losses from the 100-year wind event in the Eastern Visayas (a product of the DFCRM, scaled to match AHI), and this represents 45% of the reported US$1.4 billion in losses to Yolanda in the Eastern Visayas. The total value includes damage from precipitation flooding and storm surge (not included in this example), as well as the assets that account for the difference between AHI and nominal GRDP. 27 In response to this disaster in this location, the post-disaster program disburses a =9.4 billion), distributed uniformly among all affected house- total of US$187 million (P holds. Note that the flat tax payment mechanism used here effects a net transfer from the top quintile to the bottom four; progressive taxation, social insurance-like systems, and more complex alternatives can also be modeled. The two clusters on the right in Figure 8 show how post-disaster support would reduce wellbeing losses (without having any effect on asset losses), especially for the poorest households. The first quintile sees its wellbeing losses halved, while the impact on the richest quintile is small. In total, post-disaster support reduces wellbeing losses to US$1,265 million, a 42% decrease relative to the nominal simulation. Because post-disaster support does not impact asset losses, such programs cannot be subjected to traditional cost-benefit analyses. However, cash transfers do increase the socioeconomic resilience of the region to 50%, and the wellbeing-informed approach projects a benefit-to-cost ratio of 4.9 for this intervention. More detailed analysis would allow the costs and benefits of post-disaster support to be optimized, including realistic limitations on budgeting, targeting, and delivery.10 On an annual basis, the cost of the post-disaster program is US$472 million, equiv- alent to 32% of expected annual losses to all hazards. The program is expected to =30 billion), reduce wellbeing losses from all disasters by 17%, or US$598 million (P achieving a benefit-cost ratio of 1.3 on average. This benefit-cost ratio is lower than in the previous example because, for any individual event, the benefit-cost ratio de- pends on who is affected. When a very rich area is affected, the system may redis- tribute resources from poor (but non-disaster-affected) people to richer (but disaster- affected) people, which reduces the benefit-cost ratio. This system is provided just as an example–the cost of a post-disaster support package could be significantly re- 10 It is important to note that even if the amount of post-disaster support is equal to asset losses, it does not fully cancel wellbeing losses: indeed, post-disaster support maintains consumption, but consumption losses are larger than asset losses. This result is consistent with intuition: even if people are immediately given in cash the cost of rebuilding their houses and replacing their assets, they would still experience wellbeing losses during the reconstruction period, since assets and houses cannot be replaced instantaneously. 28 3000 Poorest quintile Second Third 2500 Fourth Disaster losses [USD per affected person] Wealthiest 2000 1500 1000 500 0 Asset loss Well-being loss Net cash benefit of Well-being loss uniform payout with uniform payout Figure 8: Asset and wellbeing losses from a 100-year hurricane (wind) event in the Eastern Visayas are shown by quintile. In the left two clusters, asset and wellbeing losses are modeled in the absence of any governmental post-disaster support. The third and fourth clusters show the net cost and expected wellbeing losses, by quintile, of a post-disaster support package in which all affected households receive a uniform payout equal to 80% of the average asset losses of the poorest quintile. Note that post disaster support reduces wellbeing losses without impacting asset losses. duced, and the wellbeing benefits significantly improved, if it were better targeted to benefit only the poorest Filipinos (e.g., by restricting beneficiaries to those in the first quintile). The impact of post-disaster support systems on disaster recovery and poverty re- duction can also be measured. Again using wind damage from a 100-year hurricane event in the Eastern Visayas as an example, Figure 9 on the following page plots the poverty gap (with only households in the bottom half of the income distribution) as reported in FIES ("pre-disaster") and immediately after the hurricane hits, with and without post-disaster support. The figure shows that a 100-year hurricane in the re- gion is expected to deepen the poverty gap by 12 percent in the first quintile, from 40 percent to 45 percent. With the post-disaster support discussed above, the magnitude 29 50% No PDS +12% Uniform PDS +6% 40% Pre-disaster Consumption poverty gap 30% +45% 20% +6% 10% +385% +172% 0% Poorest Second Third Fourth Fifth decile decile decile decile decile Figure 9: The poverty gap, defined for each income decile as the average consumption short- fall of the poverty line. Decile-level results are shown for the Eastern Visayas region as reported in FIES 2015 ("Pre-disaster") and after a 100-year wind event strikes the region. The post-disaster poverty gap is modeled both in the absence of any gov- ernmental post-disaster support ("No PDS"), and with one potential post-disaster support ("PDS") package, and is calculated only for disaster-affected households. of the shock decreases to 6%. The benefits of this program are even larger in the second and third income deciles. This paper has presented the results of a risk assessment based on an expanded framework, which includes in the analysis the ability of affected households to cope with and recover from disaster asset losses and uses “wellbeing losses” as its main measure of disaster severity. This framework adds to the three usual components 30 of a risk assessment — hazard, exposure, and vulnerability — a fourth component, socioeconomic resilience. Like the traditional components of risk management, so- cioeconomic resilience can be measured at any degree of spatial resolution, from the household to national averages. Using a new agent-based model that represents explicitly the recovery and recon- struction process at the household level, this risk assessment provides more insight into disaster risk in the Philippines than a traditional risk assessment. In particular, it shows how the regions identified as priorities for risk-management interventions differ depending on which risk metric is used. While a simple cost-benefit analysis based on asset losses would drive risk reduction investments toward the richest re- gions and areas, a focus on poverty or wellbeing rebalances the analysis and provides a different set of regional priorities. In parallel, measuring disaster impacts through poverty and wellbeing implications helps quantify the benefits of interventions that may not reduce asset losses, but do reduce their wellbeing consequences by making the population more resilient. These interventions include financial inclusion, social protection, and more generally the provision of post-disaster support to affected households. The model and data used in this analysis have many limitations. For instance, the 2015 FIES is only representative at the regional level and does not offer the ge- olocalization of households. Further, the DFCRM is itself a complex and imperfect modeling exercise. Finally, the model does not represent all coping mechanisms available to households, such as international remittances or temporary migrations. However, even with these limits and simplifications, the introduction of socioeco- nomic resilience and household characteristics into risk assessments provides useful insights and appears as a promising research agenda. 31 The authors wish to recognize the work of many colleagues who contributed to this report, including Artessa Saldivar-Sali, Lesley Jeanne Y. Cordero, Adrien Vogt-Schilb, Mook Bangalore, and the World Bank country office in Manila. They also thank the teams at the Philippines’ National Economic and Development Authority and the Philippine Statistics Authority for their contribution to the development of the model and its application to the Philippines. All errors, interpretations, and conclusions are the responsibility of the authors. [1] Sandra G. Catane, Catherine C. Abon, Ricarido M. Saturay, Edna Patricia P. Men- doza, and Krestabelle M. Futalan. Landslide-amplified flash floods—The June 2008 Panay Island flooding, Philippines. Geomorphology, 169-170:55–63, Oct. 2012. [2] Rio Yonson. Floods and Pestilence: Diseases in Philippine Urban Areas. Eco- nomics of Disasters and Climate Change, 2(2):107–135, Jul, 2018. [3] Rio Yonson, Ilan Noy, and JC Gaillard. The measurement of disaster risk: An ex- ample from tropical cyclones in the Philippines. Review of Development Economics, 22(2):736–765, May 2018. [4] Elodie Blanc, Eric Strobl, Elodie Blanc, and Eric Strobl. Assessing the Impact of Typhoons on Rice Production in the Philippines. Journal of Applied Meteorology and Climatology, 55(4):993–1007, Apr. 2016. [5] Carine J. Yi, Anawat Suppasri, Shuichi Kure, Jeremy D. Bricker, Erick Mas, Maritess Quimpo, and Mari Yasuda. Storm surge mapping of typhoon Haiyan and its impact in Tanauan, Leyte, Philippines. International Journal of Disaster Risk Reduction, 13:207–214, Sep. 2015. 32 [6] Marco G.A. Huigen and Isabella C. Jens. Socio-Economic Impact of Super Ty- phoon Harurot in San Mariano, Isabela, the Philippines. World Development, 34(12):2116–2136, Dec. 2006. [7] David Dawe, Piedad Moya, and Shiela Valencia. Institutional, policy and farmer responses to drought: El Niño events and rice in the Philippines. Disasters, 33(2):291–307, Apr. 2009. [8] Graciano P. Yumul, Nathaniel A. Cruz, Nathaniel T. Servando, and Carla B. Di- malanta. Extreme weather events and related disasters in the Philippines, 2004- 08: a sign of what climate change will mean? Disasters, 35(2):362–382, Apr. 2011. [9] Jayanta Guin and Vinita Saxena. Extreme losses from natural disasters- earthquakes, tropical cyclones and extratropical cyclones. Boston, MA: Applied Insurance Research Inc, 2000. [10] Deanna T Villacin. A review of philippine government disaster financing for recovery and reconstruction. Technical report, PIDS Discussion Paper Series, 2017. [11] Michael R. Carter and Christopher B. Barrett. The economics of poverty traps and persistent poverty: An asset-based approach. Journal of Development Studies, 42(2):178–199, 2006. [12] Stefan Dercon and Catherine Porter. Live aid revisited: Long-term impacts of the 1984 Ethiopian famine on children. Journal of the European Economic Association, 12(4):927–948, Aug. 2014. [13] Patricia Grossi and Howard Kunreuther. Catastrophe modeling: a new approach to managing risk, volume 25. Springer Science & Business Media, 2005. [14] Angus Deaton. The analysis of household surveys: a microeconometric approach to development policy. World Bank Publications, 1997. [15] Robert M Townsend. Consumption insurance: An evaluation of risk-bearing systems in low-income economies. Journal of Economic perspectives, 9(3):83–102, 1995. 33 [16] Jonathan Morduch. Income smoothing and consumption smoothing. Journal of economic perspectives, 9(3):103–114, 1995. [17] Alain de Janvry, Frederico Finan, Elisabeth Sadoulet, and Renos Vakis. Can con- ditional cash transfer programs serve as safety nets in keeping children at school and from working when exposed to shocks? Journal of Development Economics, 79(2):349–373, 2006. [18] Germán Daniel Caruso. The Legacy of Natural Disasters: The Intergenerational Impacts of 100 Years of Disasters in Latin America. Journal of Development Eco- nomics, 127(March):209–233, 2017. [19] Solomon M Hsiang and Amir S Jina. The causal effect of environmental catas- trophe on long-run economic growth: Evidence from 6,700 cyclones. Technical report, National Bureau of Economic Research, 2014. [20] Ilan Noy. The macroeconomic consequences of disasters. Journal of Development economics, 88(2):221–231, 2009. [21] Jeroen Klomp and Kay Valckx. Natural disasters and economic growth: A meta- analysis. Global Environmental Change, 26(1):183–195, 2014. [22] Mark Skidmore and Hideki Toya. Do natural disasters promote long-run growth? Economic Inquiry, 40(4):664–687, 2002. [23] Adriana Kocornik-mina, Thomas K. J. McDermott, Guy Michaels, and Ferdinand Rauch. Flooded Cities. Technical report, Centre for Economic Performance, LSE, 2016. [24] Luisito Bertinelli and Eric Strobl. Quantifying the local economic growth impact of hurricane strikes: An analysis from outer space for the caribbean. Journal of Applied Meteorology and Climatology, 52(8):1688–1697, 2013. [25] Eduardo Cavallo, Sebastian Galiani, Ilan Noy, and Juan Pantano. Catastrophic Natural Disasters and Economic Growth. Review of Economics and Statistics, 95(5):1549–1561, Dec. 2013. 34 [26] Stephane Hallegatte, Mook Bangalore, Laura Bonzanigo, Marianne Fay, Tamaro Kane, Ulf Narloch, Julie Rozenberg, David Treguer, and Adrien Vogt-Schilb. Shock Waves: Managing the Impacts of Climate Change on Poverty. World Bank Publications, 2015. [27] L. Le De, J.C. Gaillard, and W. Friesen. Remittances and disaster: a review. International Journal of Disaster Risk Reduction, 4:34–43, Jun. 2013. [28] Dean Yang and HwaJung Choi. Are Remittances Insurance? Evidence from Rainfall Shocks in the Philippines. The World Bank Economic Review, 21(2):219– 248, Jan. 2007. [29] Stephane Hallegatte, Adrien Vogt-Schilb, Mook Bangalore, and Julie Rozenberg. Unbreakable: building the resilience of the poor in the face of natural disasters. World Bank Publications, 2016. [30] Stéphane Hallegatte, Jun Rentschler, and Brian Walsh. Building Back Better: Achieving Resilience through Stronger, Faster, and More Inclusive Post-Disaster Re- construction. World Bank Group, 2018. [31] Alvina Erman, Elliot Motte, Radhika Goyal, Akosua Asare, Shinya Takamatsu, Xiaomeng Chen, Silvia Malgioglio, Alexander Skinner, Nobuo Yoshida, and Stephane Hallegatte. The road to recovery: the role of poverty in the exposure, vul- nerability and resilience to floods in Accra. The World Bank, 2018. [32] Stephane Hallegatte and Adrien Vogt-Schilb. Are Losses from Natural Disasters More Than Just Asset Losses? The Role of Capital Aggregation, Sector Interactions, and Investment Behaviors. Policy Research Working Papers. The World Bank, Nov. 2016. [33] Marc Fleurbaey and Peter J. Hammond. Interpersonally Comparable Utility. In Handbook of Utility Theory, pages 1179–1285. Springer US, Boston, MA, 2004. — 35 — This section explains the methodology and describes the model used to translate asset losses into wellbeing losses. The code of the model is freely available, and the reader is invited to refer to the code for the implementation of the principles and equations presented in this section. The household survey data cannot be made available directly, as they need to be requested from the statistical agency of the Philippines. In all applications, the model assumes a closed national economy. In terms of dis- aster risks, this means that 100% of household income is derived from assets located inside the country, and that post-disaster reconstruction costs can be distributed to non-affected taxpayers throughout the country, but not outside its borders.11 This report picks up and develops the analytical machinery of the original Unbreak- able report. Its primary innovation is in its use of the Family Income & Expenditure Survey (FIES) to disaggregate expected asset losses among representative households, resulting in a measurement of asset losses, poverty impacts, and wellbeing losses by income quintile and region in the country. While the remainder of this report of- fers a detailed description of the methodology we highlight three ways in which this iteration of the Unbreakable analysis is an extension of previous work [29, 30]: • Based on national data, the original model could not give insight into spatial heterogeneities in hazard, exposure, or asset vulnerability. This shortcoming limited the practical value of the Unbreakable framework to inform funding de- cisions at actionable level of spatial detail. The present analysis is based on hazard- and asset class-specific exceedance curves at the provincial level, even though the spatial resolution of the analysis is limited by the representativeness of the available household survey. 11 This is a serious limit in a country where international remittances have reached more than 8 percent of the gross national income in 2017, and where remittances have been shown to support post-disaster recovery [27, 28]. — 36 • The original analysis, used most recently to generate national-level indicators for 149 countries in [30], divided the population of each country into “poor" and “non-poor" groups. National averages for the characteristics of each group (e.g., income and asset vulnerability; access to early warning and financial institu- tions; social protection receipts, etc.) are used to estimate their respective asset and wellbeing losses and socio-economic resilience. Consequently, it was not possible to examine the characteristics that influence socioeconomic resilience within quintiles. In the new iteration, we can examine income and expenditures data for more insight into how best to help the poor cope with disasters. • Formerly, the model assumed exogenously that all households recover at the same pace when they are affected by a disaster. In its current iteration, the model explicitly represents disaster reconstruction dynamics at the household level using an agent-based approach in which each household acts rationally to minimize its wellbeing losses. This optimization specifies each household’s re- construction and savings expenditure rate, assuming households optimize the fraction of income they dedicate to repairing and replacing their assets. For instance, people close to the subsistence level cannot set aside much of their in- come to rebuild their assets without experiencing large wellbeing losses, and may therefore take longer to recover. In extreme cases, they may even be trapped in poverty, generating large wellbeing losses going well beyond the few years that follow a disaster [11, 12]. The model also provides a better as- sessment of wellbeing losses by distinguishing between short-lived deep con- sumption losses, and more persistent but shallower impacts. Pre-disaster situation Population & weighting The pre-disaster situation in the country is represented by the households described in the Family Income and Expenditure Survey (FIES). We use a per capita weighting — 37 (ωh ), such that summing over all households in the survey or in an administrative unit (Nh ) returns the total population (P): Nh P= ωh (1) h=0 One essential characteristic of each household is its income (ih ). As defined by the FIES, ih combines primary income and receipts from all other sources, including the imputed rental value of owner-occupied dwelling units, pensions and support, and the value of in-kind gifts and services received free of charge. The data recorded in the survey are assumed to capture the household’s permanent income, which is smoothed over fluctuations in income and occasional or one-off expenditures.12 It is important to note that the value of housing services provided by owner-occupied dwelling is included in the income data, so that the loss of a house has an impact on income (even though it would not affect actual monetary income).13 Social transfers, taxation, and remittances The enrollment and value of social transfers (isp h ) are listed in FIES, and the total cost (Csp ) of these programs to the government is given by a simple sum:14 Nh Csp = ωh isp h (2) h=0 All incomes reported in FIES are assumed to be reported net of the income tax that finances general spending of the government (for infrastructure and other services) 12 In some countries, it may be necessary or preferable to infer household income from expenditures, whether because incomes are not reported, because consumption is more stable over time, or because the official poverty statistics are calculated from consumption rather than income. 13 Similarly, the services provided by other assets (e.g., air conditioners, refrigerators) could be added as an additional income that can be threatened by natural disasters. 14 Administrative costs are not included in the assessment of the cost of the programs. When household data do not include the transfers, then transfers from social programs can be modeled on the basis of the actual disbursement rules that qualify households for participation in each program (eg, PMT score, household number of dependents or senior citizens, employment status, etc.). — 38 and of an additional flat income tax that finances social programs (rate = δtax sp ). The rate δtax sp can be estimated with the following equation: 15 Nh Nh Nh Csp ωh isp h = ωh ih = ωh ih δtax sp (3) ωh ih h=0 h=0 h=0 sp Note that, since ih includes ih , income from social programs is treated as taxable in the model. A similar approach is used to derive the tax rate (δtax pub. ) to fund post- disaster reconstruction of public assets. Remittances and transfers among households play a very important role for peo- ple’s income. Some of these transfers are within a family or a community, while others are international. The FIES provides estimates of the amount received, but it is of course impossible to represent the bilateral flows of resources among households. For this reason, remittances are modeled like an additional social protection scheme: the transfers received from friends and family are added to the social transfers, and it is assumed that these transfers come from a single fund, in which all households contribute proportionally to their income (like a flat tax). Under these assumptions, remittances can be aggregated with social protection and redistribution systems. This is of course a simplification, especially in that it does not account from international remittances, which have been shown to play a role after disaster [28]. Income, capital, & consumption Household income is equal to the sum of the social transfers and domestic and inter- national remittances, plus the value generated by a household’s effective capital stock (keff tax 16 h ), less the flat tax at the rate δsp . ih = isp tax eff h + (1 − δsp ) · kh · Πk (4) 15 Although we assume a flat tax, the model is capable of handling more complicated tax regimes, in- cluding progressive taxation. 16 Since general spending of the government is not explicitly represented in the income ih , the effective capital stock estimated here keff h is net of the resources used to finance this general spending through taxes. — 39 In practice, the household’s effective capital stock (keff h ) is estimated based on the income and transfers reported in the FIES, and the tax level δtax sp that would balance the budget: Income from assets ih − isp keff h · Πk = h (5) 1 − δtax sp Gross of taxes All household income not from transfers is assumed to be generated by household effective assets, including some assets not owned by the household (like roads and prv factories). Some of the assets represented by keff h are private (kh ), such as equipment used in family business or livestock; some assets are public (kpub h ), such as road and the power grid (and possibly the environment and natural capital); and some assets are owned by other households (koth h ), but still used to generate income by the household, such as factories. In the absence of data on different capital stocks, we use the AIR loss modeling results to calibrate the model. The AIR catastrophe model describes expected asset losses in 15 separate asset categories: all private assets are grouped together, while public assets, including transport, health, education, and utility infrastructure are prv reported individually. For each household, we distribute keff h to private (kh ) and public (kpub h ) using the fraction of private to total asset losses in the region, assuming that (1) the ratios of the different capital categories are similar for all households, and (2) the vulnerability of each household’s public assets is given by the vulnerability of its private assets. The productivity and vulnerability of these assets to various hazards can vary, so it is useful to disambiguate among them as much as the data allow. In addition, these distinctions are important to understand the consumption and wellbeing losses that follow a disaster, since the liability for reconstruction varies: households rebuild their own assets (unless they carried private insurance); the national, regional, or provincial taxpayers rebuild public assets; and other privately-held assets are reconstructed by private business owners or corporations. — 40 Precautionary savings Precautionary savings play a key role in managing disasters, but there is no esti- mate of these savings in the FIES. Also, although the FIES provides information on both income and consumption, the difference between income and consumption is highly variable and negative for many households (aggregate consumption greater than reported income), making it an uncertain indicator of savings at the household level. Instead, we calculate the average gap (income less consumption) by region and decile. We then assume that each household maintains one year’s surplus as precau- tionary savings: separate from their productive assets, and available to be spent on recovery or consumption smoothing. Household versus nominal regional GDP Based on this definition of each household’s income and assets, we note that the total capital stock (K) of any country is given by a sum over the effective capital of all households, and the portion of national GDP from household consumption (hhGDP) as reported in the FIES is given by the product of K and the average productivity of capital (Πk ). Nh K= ωh keff h (6) h=0 Table 6 on the next page lists the aggregated household income (AHI) and the nom- inal GDP (GRDP) for each region of the Philippines. Overall, the incomes reported in FIES represent 43% of the nominal GDP, subject to significant regional variations. In order to compensate for this discrepancy, and ensure that our parameters and vari- ables are consistent, we decided to work in the FIES reference. To do so, we scale expected asset losses in each region (based on the AIR catastrophe model [10]) by the ratio of the asset value calculated from the FIES to the total asset value in the AIR model. In other terms, we do not use the asset losses in PDP or US$ from the AIR model. Instead, we calculate the damage ratio or the ratio of losses to total asset val- ues in the AIR model, and then apply the damage ratio to the assets keff h calculated from the FIES data. — 41 AHI GRDP Region = [b.P/year Ratio ARMM 98.6 99.6 99.0 CAR 102.8 234.6 43.8 I - Ilocos 293.4 409.1 71.7 II - Cagayan Valley 190.2 236.8 80.3 III - Central Luzon 699.3 1,187.3 58.9 IVA - CALABARZON 957.8 2,059.5 46.5 IVB - MIMAROPA 170.1 204.8 83.0 IX - Zamboanga Peninsula 169.6 277.2 61.2 NCR 1,056.9 5,043.6 21.0 V - Bicol 237.9 282.8 84.1 VI - Western Visayas 403.5 549.8 73.4 VII - Central Visayas 388.1 867.2 44.8 VIII - Eastern Visayas 193.8 271.9 71.3 X - Northern Mindanao 221.5 517.6 42.8 XI - Davao 270.2 565.2 47.8 XII - SOCCSKSARGEN 190.9 356.0 53.6 XIII - Caraga 111.5 159.0 70.1 Total 5,756.2 13,322.0 43.2 Table 6: Aggregate household income (AHI), calculated at the regional level from the 2015 FIES, versus nominal regional productivities (GRDP). Both values are expressed in billions of US$ per year. In the following sections, we will trace the impacts of disasters on household assets and wellbeing through the following steps: 1. Disasters result in losses to households’ effective capital stock (∆keff h ). 2. The diminished asset base generates less income (∆ih ). 3. Reduced income contributes to a decrease in household consumption (∆ch ), but households affected by a disaster must further reduce their consumption to finance the repair or replacement of lost and damaged assets. 4. Household consumption losses are used to calculate wellbeing losses (∆wh ). One of the limits of the study is that we treat every event as independent, assuming that disasters affect the population as described by the 2015 FIES, and that two disas- ters never happen simultaneously (or close enough to have compounding effects). — 42 Asset losses The model starts from exceedance curves, produced by AIR Worldwide, which pro- vide the probable maximum (asset) loss (PML) for several types of natural disasters (earthquakes, tsunamis, tropical cyclones, storm surges, and fluvial and pluvial flood- ing), each administrative unit in the country, and various frequencies or return peri- ods. We make the simplification that a disaster affects only one region at a time, so that total losses in the affected region are equal to national-level losses. For each region, the input data detail the total value of assets lost due to hazards as well as the frequency of each type of disaster over a range of magnitudes. Magnitudes are expressed in terms of total asset losses (L). For example, the curves specify "An earthquake that causes at least $X million in damages in Y region is, on average, expected to occur once every Z years." When distributed at the household level, the losses L in the affected region can be expressed as follows: Nh L = Φa · K = ωh fah keff h vh (7) h=0 Setting aside for the moment the probability of a disaster’s occurrence (the "Haz- ard" component of Fig. 1 on page 5), Eq. 7 expresses total losses (L) in terms of total exposed assets (K) and the fraction of assets lost when a disaster occurs (Φa ). In the rightmost expression, losses are expressed as the product of each household’s prob- ability of being affected (fah , the "Exposure" component of Fig. 1 on page 5), total household assets (keff h ), and asset vulnerability (vh , cf. Sec. 10 on page 44). We make one important simplifying assumption, imposed by the data that are available: we assume that households are either affected or not affected; and, if they are affected, they lose a share vh of their effective capital keff h , with vh a function of household characteristics, independent of the local magnitude of the event. In case of a flood, one household is flooded or not, and if it is flooded, the fraction of capital lost will depend on the type of housing and other characteristics of the households and on a random process — but the losses do not depend on the local water depth or velocity. Similarly, an earthquake will affect a subset of the population who will experience — 43 building damages that depends on luck and the type of building — but the model does not take into account the ground motion at the location of the household. While this is of course a crude approximation, it is made necessary by the uncertainty on the exact localization of households in the FIES. With this approach, a bigger disaster is a disaster that affects more people, not a disaster that affects people more. Household exposure On the right side of Eq. 7 on the preceding page, fah is an expression of household exposure to each disaster. If we had perfect knowledge of each household’s exposure – for example, super high-resolution flood maps overlaid with the coordinates of every household (including those represented only implicitly in FIES) – we could assign a value of 0 or 1 to fah for each household and each event. Lacking this information, we interpret household exposure as the probability for any given household to be affected by the disaster when it occurs, and we assume that this probability is determined by household localization (at the highest resolution available) and characteristics. If the localization of a household in the FIES is known through the district, for instance, then the likelihood of one household to be flooded can be estimated by the fraction of the area of the district that is within the flood zone. Or, if population density maps are available and reliable, by the fraction of the population of the district living within the flood zone. This assumes that there is no relationship between income and exposure within a district. If poor people are found to be systematically more likely to be flooded, it is possible to introduce a "poverty bias" in the form of a higher probability of being affected for household with lower income. We do not have strong evidence that it is the case in the Philippines, and reviews suggest that such a bias is far from universal (see a review in [29], [31]). We therefore assume that the odds of being impacted by a given disaster are the same for all households in each administrative region. As a result, we can move fah — 44 out of the sum and drop the "h" subscript to indicate it is no longer household-specific (but it remains region-specific):17 Nh L = Φa · K = fa ωh keff h vh (8) h=0 This assumption also allows us to make a critical conceptual shift: if exposure is constant for all households in a given area, then we can reinterpret exposure (fa ) as the fraction of each household affected by a given disaster. After each disaster, of course, every household will be in exactly one of only two possible states: either it suffered direct impacts, or it escaped the disaster. On average, however, we can adopt a probabilistic approach by bifurcating each household in the FIES into two instances: affected and non-affected. We introduce this split in such a way that the total weight of each household (as well as asset losses at the household and provincial levels) remains unchanged:    ωh a = fa · ωh affected households ωh = ωha + ωhna (9)   ωh na = (1 − fa ) · ωh non-affected households Asset vulnerabilities The model assigns to each household a vulnerability (vh ), which describes the frac- tion of assets lost when a household is affected by a disaster. Again, this fraction does not depend on the local intensity of the hazard. Vulnerabilities are based on cat- egorical, qualitative information on the construction and condition of each domicile. Households are grouped into three categories: fragile, moderate, and robust, with associated vulnerabilities as described in Tab. 7 on the next page. The right-most column in Tab. 7 indicates the fraction of assets lost when a house- hold is affected by a disaster. Each category includes a smearing factor (±20% for the moderate and fragile categories, ±40% for robust dwellings). This randomness recog- nizes a degree of irreducible uncertainty, including the fact that actual losses depend not only on whether a household is affected, but also on many situational factors 17 Where higher resolution household and disaster loss data are available, it is of course possible to expand Equation 8 to the provincial or sub-regional level. — 45 FIES descriptor Category vh Strong material (galvanized iron, brick, tile, concrete, aluminum, stone, asbestos) robust 0.14 ± 0.06 Mixed but predominantly strong materials robust 0.14 ± 0.06 Light material (cogon, nipa, anahaw) moderate 0.40 ± 0.08 Mixed but predominantly light materials moderate 0.40 ± 0.08 Salvaged/makeshift materials fragile 0.70 ± 0.14 Mixed but predominantly salvaged materials fragile 0.70 ± 0.14 Not Applicable fragile 0.70 ± 0.14 Table 7: Asset vulnerability categories (e.g., for floods, water depth and velocity; for earthquakes, local soil conditions) and some random factors. For each household, a value is chosen at random within the indicated range, allowing for variation as plotted in Fig. 10 on the following page. This method of assigning asset vulnerabilities involves a critical, simplifying as- sumption: the condition of each dwelling is assumed to be a direct proxy for the vulnerability of all assets that generate income for the household. This vulnerability factor is applied not just to household (private) assets, but also the assets that a house- hold does not own, but from which it derives income (e.g., roads, utilities, factories, agriculture, and other infrastructure). In other words, the model assumes that the vulnerability of assets not owned by a household but which it still uses to generate income is well-described by the condition of their private assets—for example, that the roads used by people who live in makeshift dwellings are not paved, and equally vulnerable to being destroyed as is their home. This assumption avoids significant increases in data requirements—indeed, global data on the vulnerability of infrastruc- ture are not available—and is necessary to avoid overly-complex representations of economic interactions between each household and assets held in common. * Early warning systems When available, we incorporate data on the presence of early warning systems in affected regions. This reflects the assumption that early warning systems allow exposed households to move, reinforce, or otherwise protect their most fragile or valuable assets, thus reducing their vulnerability to disaster. Using the same assumption as in the Unbreakable report [29], we assume that households who — 46 4 10 10000 8000 3 10 HIES entries HIES entries 6000 2 10 4000 1 2000 10 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Asset vulnerability (vh) Asset vulnerability (vh) (a) Household asset vulnerability (vh ) (b) Same as (a), with y-axis plotted on a log scale Figure 10: Household asset vulnerability (vh ), constructed from qualitative wall and roof descriptions in the FIES 2015 from the Philippines. receive a warning are able to reduce their vulnerability by 20%, relative to identical households without access to early warning systems, by moving valuable items (from important papers to car or motorbikes) and implementing other mitigating measures (e.g., boarding windows, sandbagging doors). Summary of asset losses and calibration Returning to Eq. 7 on page 42, total losses in each region are defined as the losses suf- fered by each household, times that household’s likelihood of being directly affected by a disaster when it occurs. As mentioned in the previous section, losses include all assets that produce an income for the household—even those that are not owned by the household. If the vulnerability of all asset types is assumed to be linked to the vulnerability of the household’s private assets: Nh L= ωh · fa · ∆keff h (10) h=0 where: prv pub ∆keff h = v h · ( kh + kh + koth h ) (11) The calibration of fa and vh depends on the availability of hazard data. Here, we start from results from the AIR catastrophe model, which provides an estimate of L in — 47 Disaster Household capital occurs k0 prv(t = ) = 0.05× k prv kh h 0 eff Effective household capital kh Private prv asset k0 losses eff(t) = k prve t/ h kh 0 + k0 othe t/ oth + k0 pube t/ pub Other asset k0 oth losses Public pub asset k0 losses -1 t0 1 2 3 4 Time after disaster [years] Figure 11: When disasters occur, affected household reconstruct its assets at the optimal rate while staying out of subsistence, as described in Sec. 10 on page 52. Here, we illustrate the reconstruction process for a household that suffers ∆keff 0 = ∆kprv 0 + pub ∆k0 in losses at time t=to , and reconstructs with period τh = 2.1 years. each region and each possible event (different hazards, different return periods). The value of vh is based on the damage function from 10 on page 45, and is independent of the intensity of the hazards. Then, the number of affected people (or, equivalently, the probability for households to be affected) fa is calibrated such that the estimated losses are consistent with the AIR estimate.18 18 In exceptional cases where fa exceeds an upper threshold of 0.95, or 95% of households affected, the exposure is capped at 95% and the vulnerability vh is increased to match the asset losses from the AIR model. — 48 Income Losses To represent the longitudinal impacts of a disaster, it is not sufficient to consider the initial aggregate and distributional asset losses: one needs to explore the impacts on income and consumption, not only assets. Further, one needs to consider the dynamics of these impacts, not only the initial shock. The same asset losses do not cause the same effects if reconstruction and full recovery can be completed in a few months, compared with a case where various constraints make the recovery span years. To investigate this issue, asset losses are assumed from this point to be time- dependent: ∆keff eff h → ∆kh (t). When it is possible, the time variable is omitted below for simplicity and readability. Initial asset losses decrease throughout the reconstruction and recovery process as houses, infrastructure (i.e., roads and electric lines), and natural assets are repaired, replaced, and regrown. However, we assume that the income these assets had gen- erated for each household is diminished unless and until they are repaired. This includes the value households derive from their domicile; their appliances, vehicles, and livestock; and the infrastructure they use to commute to work or market. In this way, asset losses translate to income losses. Further, the reconstruction process is not free: households and governments have to invest in the reconstruction, at the expense of consumption (for households) or budget reallocation and increased taxes (for government). The objective of the model is to represent these processes, in order to estimate their longitudinal impacts on consumption, wellbeing, and poverty. It is important to note that the model assumes that households and governments aim at returning to the pre-disaster situation. It is well known that reconstruction can be used to “build back better” (for instance with more resilient building and in- frastructure, but also with more efficient and productive assets); see [30]. And the reconstruction process is sometimes transformative for an economy [32]. However, measuring the impact of a disaster as the cost of returning to the pre-disaster sit- uation is non-ambiguous and objective, making shocks comparable even when the reconstruction leads to a different end point. sp ∆ih = (1 − δtax eff sp ) · Πk · ∆kh + ∆ih (12) — 49 Post-disaster income losses are described by Eq. 12 on the previous page. The first term specifies direct losses, while the last two terms incorporate secondary and indirect impacts on household income, beyond the income losses resulting directly from a disaster. Each of these links between asset and income losses will be treated separately in this section. Direct income losses The definition of keff h includes all assets used by a household to generate income, including the value of owner-occupied dwellings that are generating “virtual” income in the form of housing services. The first term in Eq. 12 on the preceding page represents the post-tax reduction in income due to the loss of assets, assuming that the income loss is simply proportional to the asset loss. sp ∆ih = (1 − δtax eff sp ) · Πk · ∆kh +∆ih (13) Direct losses For instance, in the absence of social transfers, a pastoralist losing 3 of 10 goats would see her income reduced by 30 percent. A factory worker working in a factory losing half of its machinery would experience a 50 percent loss in income. [32] pro- vides the theoretical basis for this linear relationship, in spite of decreasing returns on capital, based on imperfect substitutability of assets. Direct income losses are par- tially offset by the social transfers tax (δtax sp ), assuming that households’ tax burden is directly proportional to their asset base. Social transfers The second term in Eq. 12 on the previous page, ∆isp h , represents the change in social transfers due to the decrease in tax revenue. As discussed above, these transfers in- clude also remittances, which are modeled as an additional social protection scheme. Households’ asset losses directly reduce their income, and as a result the tax they pay and the financial transfers they make to other households.19 Based on Eq. 3, it is easy 19 This is equivalent to assuming that the government budget is always balanced and that inter-household transfers respond instantaneously to income changes. Other changes in government spending, tax rates, and remittances are represented through the third term of the equation, ∆iPDS h , see below. — 50 to verify that the reduction in transfers is proportional to national asset losses, and these losses are fully diversified at the national level: L(t) sp ∆isp h ( t) = · ih (14) K We include explicit time dependence in Eq. 14 to indicate that social transfers re- cover to pre-disaster levels throughout the recovery and reconstruction process. Total asset losses L(t) is inclusive of all asset classes, irrespective of ownership (private, public, and other). These assets are rebuilt independently, and at different rates. Therefore, social transfers tend to recover even for households that are unable to recover their direct income losses through private asset reconstruction. Importantly, the fact that the assets used by households to generate an income have different ownerships introduces interactions across households, with each household benefiting from a rapid recovery of the others. For instance, poor people benefit from a more rapid recovery of asset-rich households, if it allows them to re-open shops and factories earlier, thereby protecting jobs and increasing income of workers. Similarly, a rapid recovery of tax payers helps governments restore social transfers and rebuild public assets. In Eq. 15, we update Eq. 12 on page 48 to reflect the structure of ∆isp h . In Eq. 16, we show that the aggregate loss in income is equal to the average productivity of capital multiplied by the total asset losses. Note that the final term, ∆iPDS h , is omitted in Eq. 16, since we have not yet discussed its funding mechanism, but costs and revenue sum to zero for all PDS systems independently. L sp ∆ih (t) = (1 − δtax eff sp ) · Πk · ∆kh + · i + ∆iPDS h (15) K h Nh Nh L sp ωh ∆ih = ωh (1 − δtax eff sp ) · Πk ∆kh + i = Πk L (16) K h h=0 h=0 — 51 Disaster c0 occurs Income Area = lost productivity of losses destroyed assets Household consumption ch Area = total value of destroyed assets Reconstruction costs Area = total value of savings + PDS Savings + PDS expenditure Household consumption -1 t0 1 2 3 4 Time after disaster [years] Figure 12: After being affected by a disaster, each household reconstructs its assets at the optimal rate, while avoiding falling below the subsistence line, as described in Sec. 10 on the following page. Here, we illustrate consumption losses through the reconstruction process for a household that suffers ∆keff0 = ∆kprv 0 + ∆kpub 0 in losses at time t=to , and reconstructs with period τh = 2.1 years. Consumption losses After their capital has been diminished by a disaster, households are able to generate less income and, therefore, can sustain a lower rate of consumption.20 Ideally, this decrease in income and consumption is not permanent, as households usually repair the damages to their dwelling, replace lost assets such as fridges and furniture, and rebuild their asset base (for instance regrowing their livestock). Because assets do not rebuild themselves, affected households will also have to forego an additional portion of their income (∆creco h ) to fund their recovery and re- 20 In addition to the loss of monetary income, this includes the loss in virtual income if the housing services provided by their home or their asset (fridge, fans, air conditioning systems) is also lost. — 52 construction.21 Total consumption losses, then, are equal to income losses plus recon- struction costs, less savings and post-disaster support (together represented by Sh ), as indicated by Eq. 17: ∆ch = ∆ih + ∆creco h − Sh (17) Total reconstruction costs are equal to the reduction in consumption needed to rebuild their asset stock, plus the increase in taxes needed for the government to rebuild public assets such as roads and water infrastructure. The contribution of re- construction costs to consumption losses at each moment depends on the ownership of the damaged assets, and on the reconstruction rate. These two dimensions will be discussed next. Consumption losses due to reconstruction costs vary by asset type (i.e., private, public, or other): 1. Affected households pay directly and entirely the replacement of the lost assets that they owned (∆kprv ). 2. All households pay indirectly and proportionally to their income for the replace- ment of lost public assets through an extraordinary tax (∆kpub ) 3. Households do not pay for the replacement of the assets they use to generate an income but do not own (such as the factory where they work; ∆koth ). Private asset reconstruction In the event of a disaster, affected households lose productive assets, which directly reduces their income. Household-level consumption losses do not end there, how- ever, as the destroyed assets do not rebuild themselves. Rather, affected households will have to increase their savings rate–that is, avoid consuming some fraction of their post-disaster income–to recover these assets. Assuming each household pur- sues an exponential asset reconstruction pathway, we calculate a reconstruction rate 21 Even though natural capital "rebuilds itself,“ people may have to reduce consumption to allow for accelerated growth. — 53 for each household that maximizes its wellbeing over the 10 years following the dis- aster while avoiding bringing consumption below the subsistence level (if possible). If the households cannot avoid having consumption below the subsistence line (for instance because consumption is below the subsistence level even without repairing and replacing lost assets), then we assume that reconstruction takes place at the pace possible with a saving rate equal to the average saving rate of people living at or below subsistence level in the Philippines (according to the FIES). To model each household’s recovery, we assume that disaster-affected households rebuild their lost assets exponentially over some number of years (τh ) after the shock, where τh specifies the number of years each household takes to recover 95% of initial asset losses. τh is related to reconstruction rate λh as follows: 1 τh = ln · λ− h 1 (18) 0.05 Given these assumptions for the response of each affected household to a disaster, the asset losses at time t after a disaster (occurring at time to ) are given by: −λh ·t ∆keff eff h → ∆kh (t) = ∆kh · e (19) In order to rebuild at this rate, the reconstruction costs to household consumption are given by: d ∆creco h ( t) = − ∆kh (t) = λh · ∆kh (t) (20) dt In the above equation, we have introduced a negative sign in order to keep this con- tribution to consumption losses positive, in accordance with our convention. To calculate for each household a reconstruction rate that maximizes its wellbeing, we plug Eq. 17 on the preceding page into the canonical definition of wellbeing: 1 1−η W= × ch − ∆ch (t) · e−ρt dt (21) 1−η — 54 Expanding these terms (and omitting social transfers and taxes for simplicity), we arrive at the following equation, where λh is the optimal reconstruction rate for each household: keff 10 1−η W= h × Π − (Π + λh ) · ve−λh t · e−ρt dt (22) 1−η t=0 This integral cannot be solved analytically, but we know that each household will ∂W maximize its wellbeing if it chooses a reconstruction rate (λh ) such that ∂λ = 0: 10 −η ∂W =0= Π − (Π + λh ) · ve−λt t(Π + λh ) − 1 · e−t(ρ+λ) dt (23) ∂λ t=0 We use this expression to determine the value of λh numerically. We note again that the optimum depends only on productivity of capital (Π), asset vulnerability (v), and future discount rate (ρ), while dependence on initial assets and absolute losses has dropped out of the expression. Fig. 13 on the next page displays the full distribution of reconstruction times τh , for two 50-year hurricanes, affecting either the NCR region or the Bicol region. In the case of the NCR region, most households can recover rapidly, in 2 to 3 years, and only a few households take more than five years to fully rebuild their asset base. In the case of Bicol, with much higher poverty rates, reconstruction is much longer, with a significant fraction of households needing more than five years to rebuild their asset base. This result is consistent with the idea of a poverty trap (see, e.g, [11]) and with the observation that poor households sometimes need a long time to get back to their pre-disaster situation [12, 17, 18]. It may also provide a theoretical explanation for the long-term consequences on incomes and growth that have been identified in the literature [19]. In addition, households must maintain consumption above a certain level to meet their essential needs. To reflect this, we use the following heuristic: if a household cannot afford to reconstruct at the optimal rate without falling into subsistence (i.e. if ih − ∆ih − λh ∆keff h < isub ), then the household reduces its consumption to t=t0 the subsistence line less the regional savings rate for households in subsistence (Rsub sav ), and uses the balance of its post-disaster income to reconstruct. Its consumption re- — 55 Event: 50-year hurricane Region : NCR Mean per cap asset loss ( k eff): US$591 30 Fraction of affected households in region [%] Mean recovery time ( ): 3.7 years 25 Region : V Bicol Mean per cap asset loss ( k eff): US$367 Mean recovery time ( ): 5.3 years 20 15 10 5 0 0 2 4 6 8 10 12 14 Optimal recovery time ( h) [years] Figure 13: Optimal reconstruction time distribution and mean for disaster-affected house- holds in NCR (red histogram) and Bicol (blue) after a 50-year hurricane event. Reconstruction time is calculated for each household after each disaster, and char- acterizes the disaster recovery pathway that minimizes wellbeing losses individu- ally for each household. mains at this level until its reconstruction rate reaches the optimum. This leads to an 1 initial reconstruction rate equal to λh = · ih − ∆ih − isub + Rsub sav . ∆kpriv h t=t0 Public asset reconstruction When disasters occur, we assume that the government borrows externally to finance the cost of public asset reconstruction, in order to speed recovery and minimize the financial burden on affected households. Eventually, the government recovers these costs through a tax, but only when recovery is complete. Through this mechanism, — 56 all households throughout the country share the cost of public asset reconstruction in the affected area. sp ωh vh kpub ih · δtax pub = Πk keff tax h (1 − δsp ) + ih · h (24) K pre−disaster income fractional public losses Note that Eq. 24 represents a distribution of the costs of rebuilding public assets to both affected and non-affected households. Like the tax to fund social protection programs, it should be understood as a universal tax with a flat rate (δtax pub ) that is by construction proportional to pre-disaster household income. Because this is conceived as a one-time tax to fund reconstruction, public asset re- construction costs are not spread across the duration of the reconstruction (in contrast to income lost due to the destruction of public assets, which does last for years after the disaster). Therefore, all of the time-dependence has been eliminated in Eq. 24, which indicates our assumption that the government does not collect the special tax at any point during recovery, but rather covers the cost of public asset reconstruction for the duration of reconstruction and collects taxes to fund this process many years later, after full recovery. Savings and post-disaster support The final term in Eq. 17 on page 52, Sh , represents households’ precautionary sav- ings, increased by post-disaster support, potentially including cash transfers to af- fected households, increases in social protection transfers, help through informal mechanisms at the community level, potential increases in remittances, and other exceptional cash transfers to households following a disaster. When available, these resources help households to smooth their consumption over time, or decrease con- sumption losses. In Section 10, we discussed one example of a post-disaster support system in which all Filipinos affected by a disaster receive a uniform payout, which is equal to 80% of the asset losses suffered by the poorest quintile. In this and other systems, the value of this PDS in included in Sh , along with any savings the household may have had before the disaster. In more realistic applications, these benefits can also accrue to non-affected households: for example, due to targeting errors in post-disaster support — 57 systems. Like the cost of public asset reconstruction, the cost of all exceptional post- disaster transfers is distributed among all households, including those in unaffected regions, long after reconstruction is complete (cf. Sec. 10 on page 55). Similarly, the rebuilding of the savings of affected households is assumed to take place far in the future, when reconstruction is complete and affected households’ incomes are back to their pre-disaster levels. Optimal consumption of savings and post-disaster support As illustrated by the gray shaded region in Fig. 12 on page 51, each household uses its savings, plus any post-disaster support it receives to smooth its consumption. More specifically, each household spends its liquid assets to establish a floor, or offset the deepest part of its consumption losses. The floor each household is able to afford is a function of the value of its income losses, savings and post-disaster support, and reconstruction rate, as well as of the average productivity of capital. Having assumed an exponential recovery with rate λh and a total value of savings Stot h , we can determine the level of this floor (γ) and the time at which the household’s savings ˆ ) by solving the following coupled equations: are exhausted (t keff h vh ˆ Stot ˆ h + γt = Π + λh 1 − e−λh t (25) λh ˆ γ = keff h Π − v h ( Π + λh ) e −λh t (26) As with the reconstruction rate, this optimization cannot be completed in a closed form without resorting to series expansions, so we combine Eqs. 25 and 26 into Eq. 27, and numerically find the value of γ that satisfies this equation: 0 = keff tot h vh Π + λh 1 − β + γln β − λh Sh (27) where β = γ · (keff h vh (Π + λh )) −1 Importantly, we assume here that the provision of PDS and savings do not accel- erate reconstruction pace, since the utilization of these resources is determined only after the rate of reconstruction is determined. This is a simplification that allows the — 58 two questions (rate of reconstruction and utilization of savings and PDS) to be solved sequentially, making it easier to solve the model. Wellbeing losses A $10 reduction in consumption affecting a rich household does not impact welfare or threaten health and wellbeing as much as the same loss would affect a poor house- hold. Welfare economics theory quantifies this difference by evaluating the utility (w) derived from a given level of consumption. Here, we use a simple constant relative risk aversion (CRRA) utility function: c1−η w= (28) 1−η The value of η, representing the elasticity of the marginal utility of consumption, is important to the modeling of the wellbeing losses; it represents both the risk aversion and the aversion to inequality in a society and is linked to preferences and values. It describes how $1 in consumption loss affects differently poor and non-poor people. Implicitly, it sets distributional weights, i.e. the weight attributed to poor people vs. the rest of the population in the aggregation of costs and benefits in an economic analysis [33]. In this study, we use a standard value of 1.5. Higher values give more importance to poor people, lead to higher estimates of wellbeing losses, and make it relatively more important to use policy instruments targeted towards poor people to reduce wellbeing risks. In order to determine the wellbeing losses that accumulate to each disaster-affected household during the reconstruction period (defined as the time for consumption to return to its pre-disaster level), we calculate wellbeing as the future-discounted time integral over 10 years after a disaster. c1 −η ∞ 1−η ho ∆ch (t) λh t ∆Wh = 1− e − 1 e−ρt dt (29) 1−η 0 coh Note that the integral evaluates to 0 when ∆ch = 0. For all other values of 0 < ∆ ch < co h , Eq. 29 has to be evaluated numerically. To balance the need for — 59 precision with computational limitations, Eq. 29 on the previous page is evaluated within the model with tmax = 10 years and dt = 1 week. tmax reconstruction c1− o η ∆ch (t) −λh t 1−η ∆Wh = h dt × 1− e − 1 e−ρt (30) 1−η coh t=0 We have assumed that the reconstitution of household savings and the taxes that fund public asset reconstruction and post-disaster support are widely distributed and far in the future, so that they reduce consumption but only after reconstruction is complete. Therefore, we assume that the wellbeing impact of using savings and PDS-related taxes can be estimated using the marginal utility of consumption of each household: long−term ∂W ∆Wh = ∆c = c−η tax tot h × ih · δpub + ∆Sh (31) ∂c Total wellbeing losses one household are equal to the sum of the loss along the reconstruction path and the long-term losses: reconstruction long−term ∆Wh = ∆Wh + ∆Wh (32) Then, the total wellbeing losses are calculated by summing over all households, using the number of individuals in the household as weight: Nh ∆W = ωh ∆Wh (33) h=0 Finally, we translate ∆W from an expression of utility back into an “equivalent consumption loss” (∆Ceq ) by determining the value of the consumption loss that an imaginary individual earning the national mean income would have to suffer in order to experience wellbeing losses equivalent to each "real" individual’s losses. This final step allows us to express wellbeing losses, like asset losses, in currency units and as a percentage of national or regional GDP. We derive ∆Ceq as follows: ∆W ∆Ceq = (34) W — 60 where: ∂W ∂ c1−η W = = = c−η avg. (35) ∂c ∂c 1 − η cavg. cavg. The result ∆Ceq is the metric we use to measure the wellbeing impact of a disaster (or of risk) on the population. It is a measure — expressed in domestic currency — of the wellbeing loss due to a disaster. If a disaster causes = P1 in wellbeing losses, it means that its wellbeing impact is equivalent to a = P1 decrease in the consumption of the average Filipino (i.e. a hypothetical individual with a consumption level equal to the average consumption in the Philippines).