Community Resilience to Cyclone Disasters in Coastal Bangladesh

Bangladesh is one of the poorest and most disaster-prone countries in the world. To address both problems simultaneously, sustainable livelihoods (SL) could be better connected with disaster risk reduction (DRR). For this purpose, one initiative implemented in Bangladesh is called the Vulnerability to Resilience (V2R) programme which ran from 2013 to 2016. This programme was primarily initiated and funded by the British Red Cross in a consortium with the Swedish Red Cross and the German Red Cross. This article presents the first evaluation of the V2R programme with three objectives. The first objective was to measure whether the selected communities have achieved community resilience characteristics as defined by the programme. The second objective was to conduct a cost-benefit analysis for the intervention. The third objective was to analyse V2R’s impact on the communities in terms of DRR and SL. Community-based focus group discussions and household-based surveys were conducted before and after the intervention (2013–2016) in two coastal communities in Patuakhali district, Nowapara and Pashurbunia. The analysis found that community members are now engaged with many hazard-resilient and vulnerability-reducing livelihood activities, using SL to implement DRR, yet these approaches were almost absent prior to V2R. Consequently, the communities have achieved resilience characteristics, being more well-organized and better connected; having better access to infrastructure, services, and economic opportunities; are more knowledgeable and healthier; and are better managing their available resources. Critiques of the concepts are discussed, although in this case, DRR based on SL has shown positive results, exactly as development theory suggests.


Bangladesh, Disaster Risk Reduction, and Sustainable Livelihoods
Bangladesh is ranked as the world's 6th most disaster-prone country [1]. From 1994 to 2013, at least 24,376 people were killed and at least 129 million people were reportedly affected by disasters Other examinations of resilient community characteristics [33,34] interrogate further many of the terms used by IFRC, indicating more nuances and subtleties. A trade-off between straightforwardness/do-ability and comprehensiveness/efficiency frequently emerges in the literature which aims to connect theory and practice for determining a community's vulnerabilities and resiliences. Whilst the simplicity of IFRC's characteristics has the potential for obscuring chronic, deep-seated vulnerabilities [12,29], this simplicity also has advantages in terms of being more easily communicated to community sectors and evaluated in the field.
Additionally, the open-endedness of these characteristics can permit evaluations with different technical levels. Characteristic 6 could invite a full, mixed methods Social Network Analysis or might merely enquire qualitatively about remittances and intra-community neighbourly assistance. Characteristic 3 could be critiqued as favouring a cash economy too much, rather than highlighting livelihoods involving bartering and subsistence-yet the latter is arguably enveloped in Characteristic 1 with economic opportunities enhancing subsistence livelihoods, thereby providing further opportunities and increased flexibility. As such, consistency in exploring the resilience characteristics of communities might not always be achieved. Conversely, when evaluating a specific intervention in a specific community, as occurs here, IFRC's characteristics provide a useful baseline which can be explored and analysed pragmatically followed by critique. In particular, IFRC's characteristics are used operationally for V2R meaning that they can be examined and evaluated in the context in which they are applied, as is done for this paper for specific case studies.
The definition of SL is adopted from Chambers and Conway, 1992 [35]: "a livelihood is sustainable when it can cope with and recover from the stress and shocks, and provide similar opportunities for the next generation".

Study Area
Cyclones regularly strike the coastal region of Bangladesh, with the vulnerabilities causing major disasters. For example, Cyclone Sidr that struck the south-west coast of Bangladesh on 15 November 2007 with wind speeds up to 240 km/h accompanied by 6 m of tidal waves caused 3406 fatalities with 1001 still missing and over 55,000 people sustaining physical injuries. Cyclone Sidr affected about 2.3 million households and damaged or destroyed 2.5 million acres of cropland, affecting around 8.9 million people experiencing estimated damages and losses of approximately US$1.7 billion [36].
Cyclone Aila hit the same region on 25 May 2009, leading to 190 deaths, injuring more than 7000 people, affecting nearly 5 million people, and destroying or damaging more than 500,000 houses. Cyclone Aila also extensively damaged more than 1400 km of embankments, an estimated 8800 km of roads, and 350,000 acres of cropland [37]. Cyclone Mahasen made landfall in the Patuakhali district on 16 May 2013, affected 1,498,579 people and left 17 people dead. More than 26,500 houses were destroyed and almost 124,000 damaged. Thousands of acres of cropland, fish ponds and fish culture were washed away by Cyclone Mahasen [38]. Cyclone Komen made landfall along the coast of south-eastern Bangladesh on 30 July 2015, but relatively small-scale damage resulted [39]. Meanwhile, sea-level rise due to climate change will be influencing meteorological hazards [14]. In total, 20.1% of Bangladesh's shoreline (i.e., 57.9 km of total coastline in the Ganges delta) is highly vulnerable to such hazards, whilst 17.5% (50 km) of the shoreline is estimated to be highly vulnerable when examining multiple hazards [40]. For this study, two communities were selected: Nowapara and Pashurbunia under the sub-district of Kalapara of Patuakhali district (Figure 1a). Annual average temperature of Patuakhali district varies from a maximum of 25.3 • C to a minimum of 12.2 • C, and average annual rainfall is 2377 mm [41,42]. Kalapara sub-district is located between 21 • 48 and 22 • 05 north in latitude and between 90 • 05 and 90 • 20 east in longitude [43]. It is bounded by Amtali Upazila (i.e., sub-district) both on the north and west, Rabnabad channel and Galachipa Upazila on the east, and Bay of Bengal on the south (Figure 1b). Meanwhile, sea-level rise due to climate change will be influencing meteorological hazards [14]. In total, 20.1% of Bangladesh's shoreline (i.e., 57.9 km of total coastline in the Ganges delta) is highly vulnerable to such hazards, whilst 17.5% (50 km) of the shoreline is estimated to be highly vulnerable when examining multiple hazards [40]. For this study, two communities were selected: Nowapara and Pashurbunia under the sub-district of Kalapara of Patuakhali district (Figure 1a). Annual average temperature of Patuakhali district varies from a maximum of 25.3 °C to a minimum of 12.2 °C, and average annual rainfall is 2377 mm [41,42]. Kalapara sub-district is located between 21°48′ and 22°05′ north in latitude and between 90°05′ and 90°20′ east in longitude [43]. It is bounded by Amtali Upazila (i.e., sub-district) both on the north and west, Rabnabad channel and Galachipa Upazila on the east, and Bay of Bengal on the south (Figure 1b).  Kalapara upazila comprises nine Unions (a Union is the smallest administrative rural geographic unit in Bangladesh). Nowapara and Pashurbunia communities are located in Lalua Union (Figure 1c). The distance from Kalapara Upazila to Lalua Union is 6.7 km; the total area of Lalua Union is 39 km 2 ; and the total population is 14,139 [44]. There are in total 13 primary and 3 secondary schools in Lalua Union. The primary occupation is fishing. It has one Union Disaster Management Committee and one Union Development Coordination Committee. The estimated population of Nowapara and Pashurbunia communities is 3800 in 900 households. The communities are located about 0-30 cm above mean sea level (MSL) [45]. This is a pilot study where Nowapara and Pashurbunia communities are selected because of their geographical location close to Bay of Bengal, economic instability (i.e., living behind poverty line) and vulnerability to cyclone disasters. These two communities are representative of the general circumstances of communities along coastal Bangladesh.
An embankment, property of the Bangladesh Water Development Board, is located within 20-500 m of the Rabnabad Channel and serves both communities (Figure 2a,b). Most households are located very close to the channel and embankment (Figure 2c,d). Almost every year during the monsoon period, the embankment breaks and the communities are inundated. Cyclones, rainfall-induced flooding, and tidal surges also frequently hit these communities. Therefore, they are affected by multiple hazards and have high vulnerability including fragile livelihoods, making them a suitable case study sites for implementing and evaluating V2R. Kalapara upazila comprises nine Unions (a Union is the smallest administrative rural geographic unit in Bangladesh). Nowapara and Pashurbunia communities are located in Lalua Union (Figure 1c). The distance from Kalapara Upazila to Lalua Union is 6.7 km; the total area of Lalua Union is 39 km 2 ; and the total population is 14,139 [44]. There are in total 13 primary and 3 secondary schools in Lalua Union. The primary occupation is fishing. It has one Union Disaster Management Committee and one Union Development Coordination Committee. The estimated population of Nowapara and Pashurbunia communities is 3800 in 900 households. The communities are located about 0-30 cm above mean sea level (MSL) [45]. This is a pilot study where Nowapara and Pashurbunia communities are selected because of their geographical location close to Bay of Bengal, economic instability (i.e., living behind poverty line) and vulnerability to cyclone disasters. These two communities are representative of the general circumstances of communities along coastal Bangladesh.
An embankment, property of the Bangladesh Water Development Board, is located within 20-500 m of the Rabnabad Channel and serves both communities (Figure 2a,b). Most households are located very close to the channel and embankment (Figure 2c,d). Almost every year during the monsoon period, the embankment breaks and the communities are inundated. Cyclones, rainfall-induced flooding, and tidal surges also frequently hit these communities. Therefore, they are affected by multiple hazards and have high vulnerability including fragile livelihoods, making them a suitable case study sites for implementing and evaluating V2R.

V2R Project Activities
V2R for these communities started in January 2013. At first, the community boundary was demarcated in consultation with the local government and civil society members. Several micro-groups were formed per community to provide information on the project, called Community Disaster Management Committees (CDMCs). CDMCs represent the focal points for all project decision-making and social development within each community, sitting for monthly meetings to review project progress and to plan for actions. CDMCs work closely with the Union Disaster Management Committee and with the Water and Sanitation (WATSAN) Committee (a formal local

V2R Project Activities
V2R for these communities started in January 2013. At first, the community boundary was demarcated in consultation with the local government and civil society members. Several micro-groups were formed per community to provide information on the project, called Community Disaster Management Committees (CDMCs). CDMCs represent the focal points for all project decision-making and social development within each community, sitting for monthly meetings to review project progress and to plan for actions. CDMCs work closely with the Union Disaster Management Committee and with the Water and Sanitation (WATSAN) Committee (a formal local Government body) and now (with one representative from each CDMC) they are part of Union Disaster Management and WATSAN Committee. The V2R project comprises three major sub-components: SL, Water, Sanitation and Hygiene (WASH), and DRR.
Under the sub-components, a range of activities was carried out including cash for work (for constructing/rebuilding community access roads); cash for starting poultry, fishery, vegetation, and grocery businesses; buying animals, sewing machines, and other livelihoods goods; establishing sanitary latrine and tube-wells; and providing safety equipment for the fishermen. People in the communities were trained for community mapping, first aid, search and rescue, and early warning. Technical and market access training was provided to all producer groups with respect to their particular specialism. Other workshops and training were conducted to enhance the capacity development of micro-groups and CDMCs and for developing social linkages with relevant NGOs and government organizations.
As part of the V2R livelihood activities, initially the poorest and most vulnerable 12% people from each community were selected for cash grants. Households identified as being landless or particularly vulnerable in other ways-such as poor female-headed households and households with people with disabilities, elderly, or children-were given priority. Each household was granted 15,000 BDT (BDT = Bangladeshi Taka, with the exchange rate being approximately 78.50 BDT = $1 USD on 5 May 2016). The expenditure was restricted according to business plans and, in some cases, priorities such as food or debt. The business plans included livelihood activities leading to producer groups such as cattle rearing (beef and milk), handicrafts (sewing and quilting), homestead gardens, aquaculture, rice production, poultry (chicken and ducks), and retail businesses (small shops). The profitability was also calculated for each business plan. A monitoring system was developed for all the beneficiary households after the grants had been disbursed.
For the cash-for-work scheme, the poorest and most vulnerable 20% of households with labour capability (including women) were selected. Each selected member was paid approximately 3000-4000 BDT for 15-20 days of work at 5 h per day. Cash-for-work included construction of internal community roads, rehabilitation of ponds used for drinking water, embankment repair, and tree plantation on embankments. In all cases, the most vulnerable households were selected in consultation with the micro-groups and CDMCs.
Overall, the SL activities comprise selecting appropriate producer groups, providing relevant training, and arranging for cash grants and/or cash-for-work. V2R particularly aimed to increase female income earning potential alongside improving household food and income sources [46].

Household and Community Surveying
At the project's beginning, a household baseline survey was conducted in early 2013. Later in February 2016, a similar survey was implemented to complete a before-intervention and after-intervention comparison. Information on household demography, education, income, savings, land ownership, profession, micro-credit, water supply, sanitation, household waste, health and hygiene, diseases, hazard exposure, disaster governance and DRR were collected using an open-ended questionnaire. The questionnaire was tested in the field before conducting the data collection. The first and fourth authors conducted the field survey with the help of other BRC field officials and local people. A total of 90 households (i.e., 10% sampling and 45 from each community) was surveyed ( Figure 3a) using a stratified random sampling method. The survey team visited the households and asked for permission from the available adult male/female head for conducting the questionnaire. Statistics from the two different years (project start and end) are compared to track the changes in community resilience and SL.
Additionally, community-based focus group discussions (FGDs) were conducted ( Figure 3b) in each community to collect generalised information about the project intervention and to get ideas on non-monetary and indirect benefits. For conducting FGDs, the CDMC members and community people were invited to exchange their views. The first author facilitated a group of attendees, introduced the topics for discussion and ensured community and representative participation. FGDs have been widely used in DRR research for generating bottom-up community information that is not always possible to analyse or explain statistically [47,48]. have been widely used in DRR research for generating bottom-up community information that is not always possible to analyse or explain statistically [47,48].

Cost-Benefit Analysis
Cost-benefit analysis (CBA) is an economic method which, despite severe critiques in the context of development [49][50][51], continues to be supported and applied in many development contexts. For example, in DRR, cost-benefit ratios have been calculated for drought interventions in Malawi [52], flood early warning in Fiji [53], and many hazards across the USA [54]. Shreve and Kelman (2014) completed a meta-analysis of dozens of cost-benefit studies for DRR demonstrating how high the benefit-to-cost ratios can be for solid DRR interventions [55]. They also summarised numerous limitations of the analyses including the absence of sensitivity checks, lack of long-term data, uncritical use of discount rates, not addressing root causes of vulnerability which are often obscured by the calculations, and failure to include any dis-benefits from DRR interventions. Additionally, most of the studies do not consider the full SL implications of DRR interventions through using baseline data from before and after the intervention. Instead, the studies focus mainly on loss and damages averted [55]. Due to the popularity of cost-benefit analyses and the credibility which decision-makers place on them, a gap remains to be filled in analysing a DRR intervention from a SL perspective. This study contributes to filling this gap by providing a detailed analysis from two communities which have baseline data from before and after an intervention, including a cost-benefit analysis.

Household Survey
Changes in household data are measured for 2013 and 2016 (the project start and end years respectively), with Appendix A providing details and major points being summarised in this section. Population growth rate is only 0.02% annually (Table A1(P1)), whereas the national population growth rate is 1.2% annually. Gender balance exists (Table A1(P1)). Everyone owns a house (100%), while some have additional land (42.2%) and land for farming (22.2%) (Table A1(P2)). The number of pupils going to school has increased by 55.2% (67-104) (Table A1(P3)).
The main earning sources of the heads of household remain almost unchanged (Table A1(P4)), but the secondary earning sources have increased dramatically (67.8%) (Table A1(P5)), demonstrating that the households engage in multiple secondary earning sources. The major occupations within the communities are fishing, farming, poultry rearing, cattle rearing and day labour (Table A1(P4)). The ownership of agricultural land has increased by 7% (Table A1(P6)).

Cost-Benefit Analysis
Cost-benefit analysis (CBA) is an economic method which, despite severe critiques in the context of development [49][50][51], continues to be supported and applied in many development contexts. For example, in DRR, cost-benefit ratios have been calculated for drought interventions in Malawi [52], flood early warning in Fiji [53], and many hazards across the USA [54]. Shreve and Kelman (2014) completed a meta-analysis of dozens of cost-benefit studies for DRR demonstrating how high the benefit-to-cost ratios can be for solid DRR interventions [55]. They also summarised numerous limitations of the analyses including the absence of sensitivity checks, lack of long-term data, uncritical use of discount rates, not addressing root causes of vulnerability which are often obscured by the calculations, and failure to include any dis-benefits from DRR interventions. Additionally, most of the studies do not consider the full SL implications of DRR interventions through using baseline data from before and after the intervention. Instead, the studies focus mainly on loss and damages averted [55]. Due to the popularity of cost-benefit analyses and the credibility which decision-makers place on them, a gap remains to be filled in analysing a DRR intervention from a SL perspective. This study contributes to filling this gap by providing a detailed analysis from two communities which have baseline data from before and after an intervention, including a cost-benefit analysis.

Household Survey
Changes in household data are measured for 2013 and 2016 (the project start and end years respectively), with Appendix A providing details and major points being summarised in this section. Population growth rate is only 0.02% annually (Table A1(P1)), whereas the national population growth rate is 1.2% annually. Gender balance exists (Table A1(P1)). Everyone owns a house (100%), while some have additional land (42.2%) and land for farming (22.2%) (Table A1(P2)). The number of pupils going to school has increased by 55.2% (67-104) (Table A1(P3)).
The main earning sources of the heads of household remain almost unchanged (Table A1(P4)), but the secondary earning sources have increased dramatically (67.8%) (Table A1(P5)), demonstrating that the households engage in multiple secondary earning sources. The major occupations within the communities are fishing, farming, poultry rearing, cattle rearing and day labour (Table A1(P4)). The ownership of agricultural land has increased by 7% (Table A1(P6)).
The individual household level income has also increased rapidly. For example, at V2R's finish, 40% of households earned 6001-9000 BDT per month, a rate which was only 8.9% at V2R's beginning (Table A1(P10)). In 2013, 31.1% of households' monthly income was under 3000 BDT; this figure is now only 2.2%. In 2013, there were no households with a monthly income in the range of 9001-15,000 but that had increased to 11.1% in 2016 (Table A1(P10)). The households are now earning more while fewer people are suffering from extreme poverty.
Household savings have increased from 3,240,000 BDT to 6,885,594 BDT on average. In 2013, no households had monthly savings in the range of 501-1000 BDT, a rate which had increased to 32.4% in 2016. An additional 13.2% of households are saving in the range of 1501-2000 BDT per month. On average, each household is saving 4050 BDT yearly (Table A1(P11)).
Household members are now well-aware of where to go for advice on issues related to livelihood support. Most of them (63.3%) now go to the government line departments for advice on agriculture, livestock, fisheries and other livelihood-related queries. This figure was only 4.4% in 2013. Previously, 51.1% people did not seek advice, a rate that has now reduced to 7.8%. Moreover, 25.6% households are now self-trained to tackle SL-related problems, compared to none in 2013 (Table A1(P12)).
Twenty-five percent of households borrow money from micro-finance institutions (MFIs) compared to 62.3% in 2013 (Table A1(P13)). The borrowers are mostly (90.9%) dependent on one MFI (Table A1(P13.1)) and they use it for investing in businesses (44.1%), family maintenance (24.7%) and to repay old debts (18.8%) (Table A1(P13.2)). The reduction in borrowing money from MFIs suggests that the households are now better off than before and have other or better sources of income for meeting basic needs.
The V2R project strategically installed tube-wells and water points within the communities. Consequently, the distance to the nearest water supply reduced by 20% in the dry season and 21% in the wet season. 82.2% of households stated that they now do not have any problem with pure drinking water supply (Table A1(P14-P18)). Now women and young girls, in particular, do not have to spend time fetching water from far away and have more time for other activities. For example, women now can dedicate time for SL activities and so increase household income.
In 2016, 87.8% of households owned a latrine and the rest used a shared latrine. In 2013, only 42.2% of households had latrines and most (43.6%) used to defecate openly (Table A1(P19 and P21)). People are now more aware of the need to clean the latrines regularly. 66.7% of households responded that they clean the latrines daily, a rate which was 0% in 2013 (Table A1(P20)). In 2016, 85.6% of households disposed of their household waste in a refuse pit. This figure was only 2.2% in 2013. Previously, they used to bury the household waste in random places (Table A1(P22)).
It is common to wash hands before eating (30.1%) in Bangladesh. After V2R, an increase was observed of washing hands after defecation (11.7%) and before cooking (8.5%) (Table A1(P23)). 93.3% of households use a soap bar and water for cleaning hands, a rate which was only 28.9% in 2013, whereas 71.1% of households usually used only water for cleaning hands (Table A1(P24)). A sharp increase is seen in awareness on cleaning hands properly and when appropriate. In addition, household knowledge and practices in safely collecting, storing and treating drinking water have led to decreases of diarrhoea and other related diseases.
The people within the community mostly used to suffer from water-borne diseases linked to frequent flooding, storm surges and less availability of pure drinking water. From 2013 to 2016, diarrhoea, pneumonia and fever reduced by 34.5%, 14.9% and 8.2%, respectively. In total, 75.8% of households confirmed in 2016 that they are not facing water-borne diseases compared to 20.4% in 2013 (Table A1(P25)). The improvement of water and sanitation services appears to have decreased the incidence of water-borne diseases. There is also a "WASH in School" component in the V2R programme. Children going to school, the rate of which has increased by 55.2%, are good agents of change for their families and surrounding households.
For health care, more people now go to a nearby general clinic (28.9%) and public hospital (8.9%). No one now visits a private doctor, whereas 22.2% did in 2013 and fewer people now go to nearby pharmacy (11.1%) as a first point of treatment (Table A1(P26)). It is expensive to visit a private doctor and pharmacies are not reliable for getting proper treatment. In this context, it is better to visit a general clinic first and then, in case of a serious illness, one can go to the public hospital. The households are following this trend, and it proves that the household members now know where to go for treatment at a reasonable cost and are acting on that knowledge.
Household members are now able to identify signs of diarrhoea which is the most significant disease in these communities. For example, most people responded that dysentery (33.0%) and vomiting (28.0%) are the two major and initial symptoms (Table A1(P27)) of diarrhoea, while in both cases, the rates were 0% in 2013. People now know how deal with diarrhoea; for example, giving Oral Rehydration Salts (ORS) packet solutions to the patient (46.6%) or preparing ORS at home (24.4%). Previously, mostly people (62.3%) used to rely on homemade ORS/fluids (namely, a sugar, salt and water solution) which is now reduced to 31.2% (Table A1(P28)). It is safer to use ready-made ORS packages to avoid contaminated water. 88.9% of respondents in 2016 described the ORS preparation process correctly (Table A1(P29)). Previously, in 2013, most respondents (90.91%) used to get the information on health and hygiene from outreach workers, but now 86.7% get the information from the Red Crescent volunteers through the CDMCs (Table A1(P30)).
From the 2016 survey, 98.9% of households are greatly concerned about disasters within the locality and 88.9% of households were affected by disasters in the past three years (Table A1(P31  and P32)). The effects of disasters were mostly related to property damage (32.6%), loss of domestic animals (26.7%), loss of income (20.0%) and damage to agricultural products (14.8%) (Table A1(P32.1)). As part of household preparedness, people can identify safe shelters for the family (32.8%). 11.8% of households stockpile necessary food, medicine and water in safer places. 25.2% of households take their animals to safer places, compared to 0% in 2013. 9.2% of households now take actions to protect their houses, a rate that was 0% in 2013. Previously, people used to think of primarily personal safety, but now they have increased concern about their household assets without a reduction in considering personal safety. In 2013, 13.0% of households used to do nothing for household disaster preparedness, a rate which has reduced to only 0.38% (Table A1(P33)). Overall, the communities now appear to be better prepared for disasters. Now, 98.9% of households get early warnings for multiple but mainly meteorological hazards. This figure was 73.3% in 2013 (Table A1(P34)). 48.2% of respondents said that they receive warnings from the newly established CDMCs. Some households also receive warnings from the Government of Bangladesh's cyclone preparedness program (16.1%), radio (15.3%) and mosques (8.8%). In 2013, most people (68.5%) would receive warnings from electronic media (i.e., radio, TV, and mobile phones) (Table A1(P35)). CDMCs appear to be a useful addition in playing an important role in disseminating warnings.
After receiving warnings, 73.3% households go to nearby cyclone shelters (17.8% more than 2013), 13.3% go to well-constructed public schools (6.7% more in 2013), and 13.3% take shelter in raised plinths (compared to 0% in 2013). Previously, 24.4% of people used to stay at home (now 0%) and 11.1% used to go to a relative's house (now 0%) (Table A1 (P36)). Respondents are much more aware of impending storms and where to take shelter.

Cost-Benefit Analysis
Costs are fixed and known for the V2R project implemented in these communities, while benefits can be harder to measure due to being both direct and indirect. CBA is one tool for estimating a project's efficiency. This section reports a generalized CBA applied to V2R for the two communities in Bangladesh. It considers only the measurable and direct costs and benefits from V2R; i.e., project activity costs, households' average yearly income, savings, household assets and agricultural product values. The benefit-cost ratio (BCR) of the V2R project is calculated as follows: The budget spent for V2R in Nowapara and Pashurbunia (2013-2016) is in Table 1.  [17]. * The project management costs were not considered. In fact, different groups of people were engaged in V2R's implementation; i.e., international monitoring inspectors, permanent staff in Dhaka and regional offices, contractual staff, and community-level volunteers. Most of them were involved in multiple projects and were in charge of several communities. This research has focused on relevant project activity costs and direct benefits. Therefore, our CBA excludes indirect costs and benefits that were not possible to measure. We have conducted a generalized CBA to evaluate the overall project activity for achieving community resilience and SL. If BCR>1, then the project is economically viable [56]. Applying a discount rate to future benefits is common in CBA. Considering various uncertainties in the future (e.g., the monetary benefits can be different with varying disaster magnitudes, changes in vulnerability, and changes in hazard parameters), the standard practice is to apply a discount rate of 12% in DRR projects [57]. In all the cases, and considering discount rates up to 20%, BCR > 1 is always found (Figure 4). In the next step, BCR is estimated over the long-term, considering a project benefit duration of 10 years, a typical value. Three different discount rates have been applied; 5%, 12% (the standard scenario for DRR [57]) and 20%. No costs are discounted, as the project cost is fixed for 2013-2016 with no costs afterwards. Assuming a 5% discount rate, BCR = 2.86 in 2026. The figure is 1.50 for a 12% discount rate. If a 20% discount rate is considered, then BCR drops below one in 2025 ( Figure 5). In all the cases analysed, apart from the extreme scenario of a 20% discount rate right at the end of the project, the V2R project measures provided higher benefits than costs incurred. Other indirect and non-monetary benefits were not possible to measure, suggesting increased chances of achieving higher BCR from the V2R project intervention.

Achieving Resilience Characteristics and SL
During the V2R project, one part of the embankment broke and the whole study area was flooded in July 2014 ( Figure 6). The people within the communities have to deal with this kind of In the next step, BCR is estimated over the long-term, considering a project benefit duration of 10 years, a typical value. Three different discount rates have been applied; 5%, 12% (the standard scenario for DRR [57]) and 20%. No costs are discounted, as the project cost is fixed for 2013-2016 with no costs afterwards. Assuming a 5% discount rate, BCR = 2.86 in 2026. The figure is 1.50 for a 12% discount rate. If a 20% discount rate is considered, then BCR drops below one in 2025 ( Figure 5). In all the cases analysed, apart from the extreme scenario of a 20% discount rate right at the end of the project, the V2R project measures provided higher benefits than costs incurred. Other indirect and non-monetary benefits were not possible to measure, suggesting increased chances of achieving higher BCR from the V2R project intervention. In the next step, BCR is estimated over the long-term, considering a project benefit duration of 10 years, a typical value. Three different discount rates have been applied; 5%, 12% (the standard scenario for DRR [57]) and 20%. No costs are discounted, as the project cost is fixed for 2013-2016 with no costs afterwards. Assuming a 5% discount rate, BCR = 2.86 in 2026. The figure is 1.50 for a 12% discount rate. If a 20% discount rate is considered, then BCR drops below one in 2025 ( Figure 5). In all the cases analysed, apart from the extreme scenario of a 20% discount rate right at the end of the project, the V2R project measures provided higher benefits than costs incurred. Other indirect and non-monetary benefits were not possible to measure, suggesting increased chances of achieving higher BCR from the V2R project intervention.

Achieving Resilience Characteristics and SL
During the V2R project, one part of the embankment broke and the whole study area was flooded in July 2014 ( Figure 6). The people within the communities have to deal with this kind of tidal flooding along with frequent cyclones and threats of sea-level rise.

Achieving Resilience Characteristics and SL
During the V2R project, one part of the embankment broke and the whole study area was flooded in July 2014 ( Figure 6). The people within the communities have to deal with this kind of tidal flooding along with frequent cyclones and threats of sea-level rise. The people in the communities rely heavily on this area for their SL activities, matching arguments that many Bangladeshis trade-off SL benefits with dangers from hazards [2]. Most of the population had been earning their livelihoods on a day-to-day basis with little scope to consider possibilities of disasters.
For example, the primary occupations of the heads of households are dominated by seasonal fishing in the sea (23.3%), working as wage labourers (37.8%), agricultural activities (17.8%) and retail businesses (7.8%). Only 1.1% work as full-time paid employees (Table A1(P4)). Due to the limited savings per household (Table A1(P11)) prior to V2R, the people's focus had been on daily tasks, leading to a potential discounting of the risks of future disasters, instead focusing on the day-to-day SL benefits [2]. Cyclones, storm surges and tidal floods are seasonal disasters with adverse impacts remaining for at least 2-3 months. Much of the communities' areas-or the entire location-is then under water, limiting sea fishing, wage labourers and retail businesses. Houses and animals can be washed out to sea while vast tracts of agricultural land can end up under saltwater, inhibiting its productivity. This level of livelihood fragility is the primary reason for persistent vulnerability within these communities.
To overcome this situation, sustainable secondary sources of household income can assist, so V2R focused on this issue. Now, the community people are more engaged in secondary livelihood activities (Figure 7), such as poultry farming and cattle rearing.
Previously, 74.4% of people were not engaged in secondary income sources, but now this proportion has reduced to 6.7%. Having plans for alternative income sources is useful for achieving household level resilience [27,33] and V2R has successfully implemented this activity for the two Bangladeshi communities. The people in the communities rely heavily on this area for their SL activities, matching arguments that many Bangladeshis trade-off SL benefits with dangers from hazards [2]. Most of the population had been earning their livelihoods on a day-to-day basis with little scope to consider possibilities of disasters.
For example, the primary occupations of the heads of households are dominated by seasonal fishing in the sea (23.3%), working as wage labourers (37.8%), agricultural activities (17.8%) and retail businesses (7.8%). Only 1.1% work as full-time paid employees (Table A1(P4)). Due to the limited savings per household (Table A1(P11)) prior to V2R, the people's focus had been on daily tasks, leading to a potential discounting of the risks of future disasters, instead focusing on the day-to-day SL benefits [2]. Cyclones, storm surges and tidal floods are seasonal disasters with adverse impacts remaining for at least 2-3 months. Much of the communities' areas-or the entire location-is then under water, limiting sea fishing, wage labourers and retail businesses. Houses and animals can be washed out to sea while vast tracts of agricultural land can end up under saltwater, inhibiting its productivity. This level of livelihood fragility is the primary reason for persistent vulnerability within these communities.
To overcome this situation, sustainable secondary sources of household income can assist, so V2R focused on this issue. Now, the community people are more engaged in secondary livelihood activities (Figure 7), such as poultry farming and cattle rearing.
Previously, 74.4% of people were not engaged in secondary income sources, but now this proportion has reduced to 6.7%. Having plans for alternative income sources is useful for achieving household level resilience [27,33] and V2R has successfully implemented this activity for the two Bangladeshi communities. The project has also provided disaster resilient tube-wells and sanitary latrines along with developing the internal road network (Figure 8a-c). After tidal flooding and cyclone disasters, water-borne diseases had been breaking out because of the habit of open defecation and due to lack of availability of pure drinking water. These two problems have now been tackled. Additionally, government interventions like constructing cyclone shelters and repairing the embankment ( Figure  8d,e) are helping people in the communities to take shelter safely on time while relocating cattle to protect livelihoods. Agricultural production is increasing because of reducing the intrusion of saline water (Figure 8e,f). The communities now have micro-groups and CDMCs and are better linked with external networks. Table 2 illustrates the changes in various indicators due to V2R and its implications for achieving the six characteristics of a safe and resilience community as defined by IFRC [25]. Each indicator is contributing to fulfilling the objectives of multiple resilience characteristics. A flowchart of V2R's community resilience and SL impacts is depicted in Figure 9.  The project has also provided disaster resilient tube-wells and sanitary latrines along with developing the internal road network (Figure 8a-c). After tidal flooding and cyclone disasters, water-borne diseases had been breaking out because of the habit of open defecation and due to lack of availability of pure drinking water. These two problems have now been tackled. Additionally, government interventions like constructing cyclone shelters and repairing the embankment (Figure 8d,e) are helping people in the communities to take shelter safely on time while relocating cattle to protect livelihoods. Agricultural production is increasing because of reducing the intrusion of saline water (Figure 8e,f). The communities now have micro-groups and CDMCs and are better linked with external networks. Table 2 illustrates the changes in various indicators due to V2R and its implications for achieving the six characteristics of a safe and resilience community as defined by IFRC [25]. Each indicator is contributing to fulfilling the objectives of multiple resilience characteristics. A flowchart of V2R's community resilience and SL impacts is depicted in Figure 9.

Resilience Characteristics-Field Validation
Just after the V2R project conclusion date on 30 April 2016, Cyclone Ruanu made landfall on the southern coast of Bangladesh on 21 May 2016. At least 30 people were killed and 700,000 people were affected as the storm surge and heavy rainfall (150-300 mm) damaged or destroyed around 80,000 houses and submerged paddy fields and standing crops [58,59].
The local people from the two study communities here reported that Cyclone Ruanu caused the water level to rise by 1.5-2.0 m. About 46 m of the embankment in Nowapara broke and caused high tide water to flow into the study area. It flooded some households and the vast agricultural field (Figure 10). The people managed to evacuate their households to the nearby cyclone shelters or other safer places by 8 p.m. on the day before the cyclone hit following warnings provided by the BDRCS.

Resilience Characteristics-Field Validation
Just after the V2R project conclusion date on 30 April 2016, Cyclone Ruanu made landfall on the southern coast of Bangladesh on 21 May 2016. At least 30 people were killed and 700,000 people were affected as the storm surge and heavy rainfall (150-300 mm) damaged or destroyed around 80,000 houses and submerged paddy fields and standing crops [58,59].
The local people from the two study communities here reported that Cyclone Ruanu caused the water level to rise by 1.5-2.0 m. About 46 m of the embankment in Nowapara broke and caused high tide water to flow into the study area. It flooded some households and the vast agricultural field ( Figure 10). The people managed to evacuate their households to the nearby cyclone shelters or other safer places by 8 p.m. on the day before the cyclone hit following warnings provided by the BDRCS. No deaths or injuries were reported and only a few thatched houses were destroyed, although most houses required some minor repairs. The communities were waterlogged for several days due to heavy rainfall and flooding from the Rabnabad Channel. The high tide did not wash away the sanitary latrines or damage the tube-wells, nor did it cause harm to cattle and poultry. No deaths or injuries were reported and only a few thatched houses were destroyed, although most houses required some minor repairs. The communities were waterlogged for several days due to heavy rainfall and flooding from the Rabnabad Channel. The high tide did not wash away the sanitary latrines or damage the tube-wells, nor did it cause harm to cattle and poultry. After a few days, people returned to their houses and continued their livelihoods. The ample supply of drinking water continued, as did the usability of the sanitary latrines. As the agricultural fields were inundated by saltwater, the farmers and day labourers are not yet able to continue cultivating the land. They are instead shifting livelihoods to fishing, as this is high time (rainy season) for sea fishing. To recover from the damage, they are focusing on repairing the embankment and damaged houses while providing support for growing rain-fed lowland rice in tidal saline water.
In 1970 and 1991, cyclones of similar strength killed more than 300,000 and 100,000 people, respectively, in Bangladesh. Then, over the last decade, two cyclones of similar strength, Sidr and Aila, hit Bangladesh's coast, each killing two orders of magnitude fewer people than the earlier storms, many of whom were fishermen out at sea who did not return to land in time [59]. Damage and losses from Cyclone Ruanu were substantially less than these earlier events. It is argued that this dramatic reduction in loss of human lives was possible due to an extensive DRR programme across the coastal region of Bangladesh [59] to which V2R contributed amongst the myriad of local endeavours along the coast.
The two communities in this study have shown strong resilience characteristics through V2R supporting SL. The community people including the fishers evacuated safely before the cyclone struck, an action possible because of being highly connected to external organizations for warnings while being internally cohesive for coordinated, effective response. They received warnings in time, had completed associated training evacuating along safer routes to shelter, knew when to come back from sea while fishing, and were able to manage their properties before and after the evacuation. After a few days, people returned to their houses and continued their livelihoods. The ample supply of drinking water continued, as did the usability of the sanitary latrines. As the agricultural fields were inundated by saltwater, the farmers and day labourers are not yet able to continue cultivating the land. They are instead shifting livelihoods to fishing, as this is high time (rainy season) for sea fishing. To recover from the damage, they are focusing on repairing the embankment and damaged houses while providing support for growing rain-fed lowland rice in tidal saline water.
In 1970 and 1991, cyclones of similar strength killed more than 300,000 and 100,000 people, respectively, in Bangladesh. Then, over the last decade, two cyclones of similar strength, Sidr and Aila, hit Bangladesh's coast, each killing two orders of magnitude fewer people than the earlier storms, many of whom were fishermen out at sea who did not return to land in time [59]. Damage and losses from Cyclone Ruanu were substantially less than these earlier events. It is argued that this dramatic reduction in loss of human lives was possible due to an extensive DRR programme across the coastal region of Bangladesh [59] to which V2R contributed amongst the myriad of local endeavours along the coast.
The two communities in this study have shown strong resilience characteristics through V2R supporting SL. The community people including the fishers evacuated safely before the cyclone struck, an action possible because of being highly connected to external organizations for warnings while being internally cohesive for coordinated, effective response. They received warnings in time, had completed associated training evacuating along safer routes to shelter, knew when to come back from sea while fishing, and were able to manage their properties before and after the evacuation.
Due to V2R, after the cyclone, the communities could access safe drinking water, sanitary latrines, and other community necessities. All the households now have secondary sources of income, so some temporarily shifted livelihoods while rebuilding their primary income sources. In addition, it is expected that rice cultivation will soon start again, as a crop that can withstand the salinity and monsoon along the coast is available. V2R was tested soon after it finished, demonstrating community resilience and SL through connecting with DRR.

Critical Reflection
This article covers only community resilience as defined by the V2R implementers, IFRC [25]. As per Figure 11, resilience functions at multiple levels [25] and various traits and definitions are not always fully accepted or explored in research or practice [11,24,27,60]. Local government or policies undertaken by the national government can influence resilience at individual, household and community levels. For example, delays in embankment re-building can seriously affect community resilience and community resilience building. As well, the Government of Bangladesh plans to start constructing a new seaport (the Paira Bandar seaport in Kalapara Upazila) at Rabnabad Channel. Therefore, some people in the two studied communities might experience forced eviction which would significantly change their livelihood patterns and undermine the work completed by V2R. Due to V2R, after the cyclone, the communities could access safe drinking water, sanitary latrines, and other community necessities. All the households now have secondary sources of income, so some temporarily shifted livelihoods while rebuilding their primary income sources. In addition, it is expected that rice cultivation will soon start again, as a crop that can withstand the salinity and monsoon along the coast is available. V2R was tested soon after it finished, demonstrating community resilience and SL through connecting with DRR.

Critical Reflection
This article covers only community resilience as defined by the V2R implementers, IFRC [25]. As per Figure 11, resilience functions at multiple levels [25] and various traits and definitions are not always fully accepted or explored in research or practice [11,24,27,60]. Local government or policies undertaken by the national government can influence resilience at individual, household and community levels. For example, delays in embankment re-building can seriously affect community resilience and community resilience building. As well, the Government of Bangladesh plans to start constructing a new seaport (the Paira Bandar seaport in Kalapara Upazila) at Rabnabad Channel. Therefore, some people in the two studied communities might experience forced eviction which would significantly change their livelihood patterns and undermine the work completed by V2R. These multi-scalar processes and changes over which the communities have little control are a major limitation of V2R and IFRC's resilient community characteristics. Although far from being exclusively created externally, vulnerabilities are often imposed from outside [11][12][13] with little notice and little opportunity to counter them. Conversely, resiliences are not exclusively an intra-community matter, because national and international support systems are frequently assumed to exist and to be an inevitable part of a community dealing with disaster, such as remittances [61] and national social safety nets. V2R and IFRC's resilient community characteristics will not necessarily capture all such dimensions, depending how extensive the work is.
Meanwhile, other hazards and hazard drivers such as climate change, sea-level rise and tsunamis-along with social factors such as geopolitical unrest, economic recession, discrimination, inequality and corruption-continue to affect local vulnerabilities and resiliences. Although relocation is generally not preferred by the inhabitants, coastal Bangladesh will be seriously affected by climate change over the next decades, with elevated air and sea temperatures, sea-level rise, and more intense cyclones all threatening livelihoods and lives [40,62,63]. Meanwhile, a devastating earthquake could easily strike across Bangladesh at any time, even before sea-level rise's main effects are witnessed [64,65]. In addition, population movements, geopolitical unrest, and These multi-scalar processes and changes over which the communities have little control are a major limitation of V2R and IFRC's resilient community characteristics. Although far from being exclusively created externally, vulnerabilities are often imposed from outside [11][12][13] with little notice and little opportunity to counter them. Conversely, resiliences are not exclusively an intra-community matter, because national and international support systems are frequently assumed to exist and to be an inevitable part of a community dealing with disaster, such as remittances [61] and national social safety nets. V2R and IFRC's resilient community characteristics will not necessarily capture all such dimensions, depending how extensive the work is.
Meanwhile, other hazards and hazard drivers such as climate change, sea-level rise and tsunamis-along with social factors such as geopolitical unrest, economic recession, discrimination, inequality and corruption-continue to affect local vulnerabilities and resiliences. Although relocation is generally not preferred by the inhabitants, coastal Bangladesh will be seriously affected by climate change over the next decades, with elevated air and sea temperatures, sea-level rise, and more intense cyclones all threatening livelihoods and lives [40,62,63]. Meanwhile, a devastating earthquake could easily strike across Bangladesh at any time, even before sea-level rise's main effects are witnessed [64,65]. In addition, population movements, geopolitical unrest, and inadequate governance have the potential to trigger various urban disasters [66][67][68]. Therefore, achieving community resilience through SL in rural Bangladesh can be treated as a tool for urban governance and urban resilience as well.
In fact, achieving resilience at one level does not necessarily reduce vulnerability at other levels or overall. Building resilience and reducing vulnerability are iterative processes which cross scales [11,12]. The complementarity of these actions was reflected in V2R and in the analyses, showing that vulnerability and resilience are not construed as being opposites in this work, instead expressing overlapping but different tracks of a community aiming to deal with disaster and disaster risk. Thus, the case studies and analysis here provide further empirical evidence and operational experience to overturn the viewpoint that vulnerability and resilience are opposite sides of the same coin. Despite the high vulnerabilities and multiple hazards along the coastal regions of Bangladesh, these locations also provide ample SL opportunities and demonstrate substantial resiliences [69,70]. This mixed situation suggests that focusing at the local level only could not account for the complete picture of the community's vulnerabilities and resiliences. Failure to achieve resilience at national and global levels ( Figure 11) would seriously hamper activities and projects at the community level, such as V2R.
V2R and the IFRC characteristics of a resilience community do not explicitly indicate at which scales efforts should most be focused or how to balance the cross-scalar characteristics of vulnerability and resilience. V2R does provide a solid baseline for understanding local contexts to ensure that at least the local scales of vulnerability and resilience are understood and addressed. Scope exists to apply V2R at other scales and to connect scales, perhaps then realising the potential of many suggestions regarding resilience which aim to close the theory-practice gap [24,[27][28][29].

Conclusions
The aim of this article was to evaluate the implementation of V2R in two coastal communities: Nowapara and Pashurbunia in Patuakhali district, Bangladesh. The three objectives of this study were met. The first objective was to measure whether the selected communities have achieved the community characteristics set out by V2R. The communities achieved the six characteristics of resilient communities as defined by IFRC [25]. The community is now well-connected with external entities, they have better infrastructure services (e.g., access to sanitation and water supply), they are economically solvent, they are working as a team (representing social cohesiveness), they are trained and have adequate knowledge to tackle the upcoming disasters and they are managing the natural assets. The second objective was to conduct a cost-benefit analysis for the V2R intervention. In almost all the scenarios, the BCR was calculated to be greater than 1. It was estimated that the V2R project would generate benefits to the communities for at least the coming decade without any serious disruption (considering the 12% discount rate which has been standard for DRR interventions, but a conclusion also emerging for a wide range of discount rates).
The third objective was to analyse V2R's impact on the communities in terms of DRR and SL. The people in the communities are now engaged with secondary livelihood activities (e.g., poultry and livestock rearing, retail business, courtyard plants, and selling home-made cloths) alongside the primary occupations (i.e., sea fishing and farming). This trend was almost absent in the past. Here, V2R used SL as the primary tool for achieving resilience in the communities, prompting DRR based on SL-exactly as the theory suggests [12,13,70].
DRR is long-term process and it is not possible to eliminate disaster risks entirely. Yet DRR can be addressed and achieved gradually, particularly through SL. By securing community SL that are not easily disrupted or damaged by various hazards, by having access to diverse resources and connections, and by ensuring proper natural resource management, DRR and SL can be linked through IFRC's characteristics of a resilient community.
Author Contributions: All authors contributed to design, method, and analysis reported in the paper. Ahmed and Saha conducted the fieldwork, collecting and processing the data reported in the paper. Ahmed conducted the analysis and led the writing of the manuscript. Ahmed developed and prepared all tables and figures. All authors commented on the manuscript providing insights used in the analysis and discussion.

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript:    Great concern 98.9 Little concern 1.1 No concern 0.0 I don't know/no answer 0.0