Flood risk reduction and climate change adaptation of manufacturing firms: Global knowledge gaps and lessons from Ho Chi Minh City

Flooding places continuous stress on small-and medium-sized enterprises (SMEs) particularly in developing and transition economies that depend on their firms ’ performance, but may have not fully developed flood protection infrastructure. Yet, detailed knowledge about whether and how firms take adaptation action against flood hazards, including potential adaptation barriers, is surprisingly thin. We respond to this gap by offering an empirical analysis of small-and medium-sized manufacturing firms in Ho Chi Minh City (HCMC), one of the front lines of future environmental risk. Drawing on qualitative interviews, our study has two main aims: first, to understand if adaptive action occurs more intensively among firms that previously faced high direct impacts, and second, to shed light on other internal firm characteristics as well as external conditions that determine firm decisions to undertake flood adaptation measures. We find that the majority of firms cope reactively to prevent severe flooding effects. Interestingly, experience with past events, including those of high impact, do not directly lead to long-term strategic adaptation. A lack of business capabilities and financial capacity, combined with insufficient support systems, largely hamper proactive adaptation. This leads to increased risks and self-reinforcing effects, which quickly endanger the future business viability of firms. Future research on adaptation to climate change should seek to improve understanding of adaptive capacities among SMEs, which are quite different from those of large international corporations.


Introduction
Disasters and climate change risks are a growing concern for business leaders around the world.According to the 2020 edition of the World Economic Forum's (WEF) annual Global Risks Report drawing on feedback from business leaders and other high-level decision-makers, the top three global risks in terms of likelihood and impact are extreme whether events, the failure of climate action for mitigation and adaptation, and natural disasters [1].Pre-Covid the urgency of environmental risks had thus clearly overtaken business leaders' other concerns, including economic, geopolitical, technological, and societal concerns.
Yet, despite this risk assessment by top-level business leaders, detailed knowledge about the environmental and climate risks faced by the private sector is surprisingly thin.The 2013 Global Assessment Report on Disaster Risk Reduction, which focused on economic risks and private sector impacts from disasters, draws a conceptual distinction between direct losses (i.e., destruction of assets and stocks), indirect losses (i.e., business interruptions due to direct losses or supply chain interruptions), wider impacts (i.e., other consequences such as loss of market share or clients), and macroeconomic effects (i.e., all of the above including their effects on business performance) [2].However, estimates on disaster losses are scarce and characterized by high uncertainty, even with respect to previous direct and indirect losses, to say nothing of future losses.Estimates of the costs of climate change by the end of the 21st century range from only a few percentage points of GDP reduction to a 25% decrease in global per capita output under a 3 • C warming scenario [3].However, apart from macro-economic modelling, damage or loss projections at the firm leveland by different types of firmsare in short supply.Data is particularly thin for developing countries and transition economies, as well as for small-and medium-sized firms (SMEs), by contrast with data for large corporations.
Even less is known about the ways in which corporations can or do adapt to these risks, particularly with an eye toward upcoming climate change risks [4].While the rapidly growing body of climate change adaptation literature has focused strongly on the adaptation of entities such as private households, state organizations, or small-scale farmers, firms and especially SMEs offside the big global corporations tend to be neglected by the adaptation literature [4].This is a surprising and pressing gap given that private investments contribute the lion's share of overall investments in most economiesthus constituting the majority of assets exposed to future disasters -, while corporations comprise major parts of the economy.Studies looking into the reporting of 1600 adaptation strategies of corporations which are now requested to disclose their climate activities under the new Climate-related Financial Disclosures (TCFD) mechanism found severe blind spots within the firms' current treatment of climate risks and adaptation [5].Moreover, the IPCC's Fifth Assessment Report revealed a strong gap in the existing scientific knowledge on firm-level adaptation to climate change hazards [6].
This paper helps to close this gap.To understand whether and how firms are adapting to natural hazards and which barriers may persist, we offer an empirical analysis on one of the front lines of environmental risk and adaptation: we focus on manufacturing firms in Ho Chi Minh City (HCMC), an increasingly flood-prone city that is not only highly exposed to current flooding, but also to intensifying climate change hazards in the future.Meanwhile, HCMC and urban Vietnam as a whole are also undergoing substantial changes in terms of institutional and socioeconomic framework conditions for adaptation [7,8].We focus on manufacturing firms because, by contrast with the service sector or other companies, this sector is among those that face the highest potential impacts, both in terms of direct and indirect losses.Flooding that manufacturing firms face in their production facilities can cause damage to materials, products and machinery, as well as give rise to logistical problems, such as delivery delays and disrupted supplies.This is due to these firms' localized and immovable production assets, as well as their high dependence on supply chains.A focus on HCMC further allows us to closely consider one of the most hazard-prone regions of the world.According to the WEF's Report on Regional Risks for Doing Business [9], natural disasters are seen as the single most important business risk in East Asia and the Pacific region.This aligns with other studies, such as the World Risk Report [10], the INFORM Report [11] and the ND-Gain [12].Highlighting empirical lessons from HCMC offers important insights for firms in other cities of the region and eventually globally.
Specifically, we explore the extent to which manufacturing firms prepare themselves against flood impacts.We are particularly interested in understanding if adaptive action occurs more intensively among firms that previously experienced high direct impacts, such as the destruction of assets and stocks, and which other internal firm characteristicsor even external conditions, such as the institutional supportshape firms decisions to undertake flood adaptation measures.
The article is structured in six parts.Section 2 develops a conceptual framework for the analysis based on slowly emerging literature on firm adaptation.Sections 3 and 4 introduce our study area and methodology.
Section 5 presents and discusses our empirical findings on manufacturing firms' adaptation strategies in HCMC.In section 6, we discuss the implications of our findings for future research.

Conceptual considerations: firms' flood response between action and inaction
Over the last two decades, adaptation research has contributed significantly to improving our understanding of responses to natural hazards such as flooding (e.g., Refs.[13][14][15][16][17].However, existing research tends to exclude business perspectives and responses, particularly in developing countries and transition economies [18].Although the private sector has always greatly suffered from climate disasters [2], its responses were often not tracked or featured in great detail in global reports on climate change assessments (e.g., IPCC Assessment Reports prior to 2014).Research on firm adaptation to natural hazards and climate change remains limited, but is slowly rising.Especially since the IPCC Fifth Assessment Report in 2014, research into adaptive firm behavior and strategies is gaining increasing attention.The emerging body of literature focuses on current on-ground adaptation practices in social-ecological systems [17].In such systems, adaptation refers to "the process of adjustment to actual or expected climate and its effects, in order to moderate harm or exploit beneficial opportunities" [19]; p. 678).With respect to empirical research on firm adaptation strategies, Linnenluecke et al. [20] suggest that most studies focus on changing business conditions, but neglect how firms adapt to natural environment dynamics, including adaptation measures that firms undertake in anticipation of future hazards attributed to climate change.Daddi et al. [21] draw the same conclusions and argue that theoretical frameworks on firm adaptation strategies are still lacking, although management studies place research into environmental change on their scientific agenda years ago.
Recently, Neise and Revilla Diez [22] derived a typology of how manufacturing firms strategically adapt to changes in their natural environment including flood risk (see Table 1 for a definition), focusing on Jakarta, which is a similar setting to HCMC.They distinguished between proactive, relocation, reactive, surrendering, depending, and collaborative strategies, which can be tracked dynamically and are often overlapping.Proactive strategies are defined as continuous and aimed at

Flood Risk
The combination of the probability and expected amount of loss and damage resulting from one or more flood hazards (e.g., fluvial and pluvial, or compound flooding sources).
Drawing on [16,23,24] (for a discussion on risk definitions, see [25] Flood Exposure The presence of people, infrastructure, housing, production capacities, and other tangible human assets that could be adversely affected by flood hazards.
Drawing on [19,26] Adaptation In human systems, the process of adjustment to actual or expected climate and its effects in order to moderate harm or exploit beneficial opportunities. [19] Direct impact (direct losses) The destruction of assets and stocks (e.g, damage of buildings, machineries).
Drawing on [2,27] Wider impact Longer-terms social and economic effects (e.g., in productivity or in the macro economy).
Drawing on [27] R. Leitold et al. long-term behavioral shifts, a "process of mediating between sustaining and re-organizing" [28]; p. 29).If firms realize that in situ adaptation is not sufficient, the most drastically adaptive measure would be to relocate production sites for future production lines.Short-term, reactive strategies in response to specific flooding events, in turn, are often understood as 'coping' [28] or 'resistance' [29].Firms focus on avoiding flood impacts and securing resources and products [30].Collaborative strategies rely on cooperation with other companies, the community, or the state, whereas depending and surrendering strategies are characterized by firm inactivity.In the former, firms count on large-scale adaptation measures (e.g., dike-systems) initiated by state authorities, such as industrial parks (e.g., local drainage systems), business networks, or NGOs [22]).In the latter, firms follow a wait-and-see strategy [31] or even surrender.
Based on this conceptualization, we argue that firms' adaptive behaviors occur along a continuum between action and inaction, often with hybrid strategies.It remains unclear what efforts firms put into adaptation measures, and why firms engageor do not engagein proactive strategies, reactive coping mechanisms, or even remain inactive.We assume that various factors at different scales influence flood adaptation strategies among individual manufacturing firms.In this study, we derive these influencing factors from available empirical evidence in order to further develop this emerging conceptual framework.
First, we expect that firm adaptive behavior in flood-exposed areas depends heavily on their previous flood experience including the degree of flood impacts [32,33].Recent research to assess the impact of floods on manufacturing firms has been growing.Some empirical research investigated short-and long-term impacts, and specifically targeted the assessment of losses and damages due to natural hazards [34,35].Firms in the manufacturing sector suffer the most in terms of losses and damage caused by flooding, as they involve hard-to-change infrastructure (i.e., heavy input material and machines for production processes) and are characterized by strong dependence on the upstream supplier system.For SMEs, in particular, flooded production sites pose serious risks to their viability, as flooding critically affects their ability to continue business operations over time [36].They suffer from direct impacts like property damage, spoiled stocks, and electricity shortages.These probably interrupt normal operations and thus lead to additional indirect impacts such as delayed customer deliveries which are usually not covered by insurance [35].Moreover, during a flood event, some employees are either unavailable as some roads become impassable, or they have to take care of the water inflow and cannot perform their normal work routines [36].Agrawala et al. [32]; Kato and Charoenrat [37]; and Linnenluecke et al. [38] emphasize that flood-experience is a key driver for firms' adaptive behavior and investment decisions.However, even if firms must adapt adequately to floods in order to remain successful, it is not clear to what extent high flood exposure (see Table 1 for a definition) leads to the implementation of proactive adaptation [39].
Second, in line with Averchenkova et al. [4]; Gasbarro and Pinkse [40]; Halkos et al. [41]; and Howe [42]; we assume that firm adaptive behavior in flood-exposed areas is also influenced by internal firm characteristics.Empirical research on firms' internal resources and dynamic capabilities for flood adaptation in East and Southeast Asian contexts is growing [22,37,41,[43][44][45].Trinh and Thanh [46] emphasize that insufficient technological capabilities and insufficient access to credit prevent firms from implementing effective adaptation.In such cases, firms are either pushed to take advantage of temporary opportunities or else to surrender.The implementation of long-term, planned adaptation measures such as business continuity management, property-level flood protection, or relocation plans thus strongly depends on financial liquidity [35,47].The organizational learning approach [31] assumes that firms rely on their dynamic capabilities, including higher order competencies such as skills and knowledge, in order to autonomously adapt to changing business situations [44,48].Along these lines, proactive adaptation is undertaken in an anticipatory manner, which is realized at the level of managerial risk perceptions and preparations to reduce impacts of future flood events [42,49,50].Moreover, the managerial perception of responsibility influences whether a firm takes adaptive action by itself or transfers responsibility to the next available level of authority, for example to industrial park management or local district authorities.
Third, however, Tompkins et al. [51]; Meinel and Schüle [49]; and Wachinger et al. [52] argue that even if firms are highly impacted by flooding and firm managers have the knowledge, capacity, and resources to undertake measures (i.e., the internal factors), this does not automatically lead to strategic adaptive action.Other external factors in the business environment are therefore at play as well.Yet, the role of external factors and processes for influencing firms' adaptive decision-making to recurring flood events still remains poorly understood [17,30,53].Initial insights suggest that the role of the state in flood risk management and regulatory drivers appears to be crucial for individual flood adaptation [4,32,44].Halkos et al. [41]; for instance, show that institutional conditions, governmental support and guidance have direct positive impacts on the measures that SMEs undertake against the negative consequences of extreme weather events.Left by themselves to adapt to floods, SMEs particularly struggling to survive, as they are exposed to many other risks (e.g., economic crises, skills shortages, competition).Furthermore, Averchenkova et al. [4] and Pulver and Benney [54] indicate that foreign ownership as well as an export orientation are drivers for organizational business adaptation.
Against this backdrop, this article contributes to an emerging research focused on the study of firm adaptation.We empirically investigate firms' responses to recurrent flood events and expected future impacts of climate change, applying an actor-centered perspective.We hope to provide external decision makers and administrative bodies with new insights into firm level decision-making processes.

Study area: flood risk in HCMC
HCMC offers key lessons for other expanding flood-prone areas in Southeast Asia and beyond.The city is already affected by frequent flood events, which are expected to be exacerbated by climate change in the future.Being home to close to 9 million inhabitants, flood risks in HCMC can be attributed to a combination of heavy rain (particularly during monsoon seasons), discharge from upstream reservoirs, and storm surges that often coincide with high tide and annual rainfall peak [55,56].
The city is located in South Vietnam on the north-eastern edge of the Mekong Delta, downstream from the Saigon-Dong Nai River.The Saigon River Basin is a complex hydrological system, characterized by about 8000 km of interconnected rivers and canals, extensive low-lying marshlands, and seasonally extreme rainfalls [55,57,58].Usually, 90% of the annual rainfall (about 2000 mm in total) falls during the rainy season, from May to November [59].In addition, the semi-diurnal tide influences the urban water system, causing periodic rises in water levels.As about 60% of the city sits at a maximum elevation of 1.5 m above mean sea level, sea level rise (SLR) due to climate change has become one of the most pressing issues in HCMC region [60,61].Sea level on the coast has already risen by about 20 cm over the last 50 years [62].SLR projections in HCMC area are slighly higher than the global mean [63].A recent study by Scussolini et al. [56] modelled a 100 year flood in the present, which flooded 2,3% of HCMC with at least 1 m of water.By 2100, this increases to 11% in a high emissions scenario (RCP 8.5) and 28% in a high-end emissions scenario (RCP 8.5 high-end).For SLR in high emissions scenarios, they use central values presented in Jevrejeva et al. [63]; specific for the region of the East Sea.
Currently, greater HCMC generates about 40% of the country's GDP [64].Since the liberalization policies in 1986 (Doi Moi), the constant economic growth of the Southeast region of Vietnam is driven by massive foreign direct investment (FDI) flows, and thus supported by the emergence of the private sector.About 350,000 private firms operate across a range of sectors.The highest share in the total number of manufacturing firms in 2017 is accounted by the garment industry (13%), followed by the production from metal (12%), the production from rubber and plastic (9%), and the production of food and beverages (9%) [65][66][67].SMEs account for almost 96% of all manufacturing firms, and employ 30% of all manufacturing employees in HCMC [65].
HCMC's manufacturing labor market is characterized by young employees who often migrate from rural areas to labor-intensive industrial areas [68].Although the SME sector is an economic key driver in HCMC, representing the backbone of the Vietnamese economy, the institutional setting is biased in favor of FDI and multinational firms [66,69].SMEs generally encounter an ineffective government support and a lack of financial accessibility, which are considered the main barriers to SME growth in Vietnam [70].
Additionally, from a business perspective, flooding could have severe impacts, especially threatening the viability of SMEs (see below).Manufacturing firms located in HCMC's urban area are sensitive to compound flooding sources: Fluvial flooding (when river water levels rise due to excessive rain or periodical high tides) and pluvial flooding (i.e. flash floods and surface water).Those firms located close to the Saigon river are mainly exposed to fluvial flooding, which is aggravated by upstream water reservoir releases [64].A recent study by Leitold and Revilla Diez [66] shows that the highest proportion of manufacturing firms exposed to future SLR in HCMC comes from the production of prefabricated metal parts, rubber and plastic, and the garment industry.Firms in the western parts of the city are mainly exposed to urban flash and pluvial flooding following high intensity or prolonged rainfall, which lead to overwhelmed urban drainage systems and clogged waterways [55].Constant land subsidence also exacerbates the flooding situation [71].
In the French colonial period, the urban core of HCMC was built on higher areas (4-6 m above mean sea level), presently located in the urban center of District 1 (see Fig. 1) [64,69].HCMC has been expanding rapidly.The population increased from 3.5 million in 1976 to almost 8.3 million people in 2016 [72].Urban development pressure and strong economic development led to a shortage of space and pushed newly developed manufacturing firms to lower elevation areas beyond the city center.As a result, contemporary industrial areas face high flood exposure [66]).
Duy et al. [64] suggest that economic planning in HCMC does not sufficiently consider topographic conditions or urban expansion trends.As stated by the World Bank [73], Vietnam has lost between 1 and 1.5% of GDP annually over the last two decades due to water-related natural disasters.This number is predicted to sharply rise due to the effects of climate change, negatively impacting the economic powerhouse of the country.Some efforts have been recently made to implement flood control strategies, such as the recent Flood Prevention Planning 1547, approved in 2015 by the Ministry of Agriculture, whose mandate also includes disaster risk reducion.This project aims to implement an integrated flood risk management approach, including tidal sluice gates and embankments east of the Saigon River.However, there is still a discrepancy between current flood risk planning, the implementation of measures, and their effectiveness for different social and economic realities.For example, planning to improve canal bank revetments or storm water sewage systems mainly address FDI-based and private firms in industrial zones [74].Thus, firms outside of industrial parks especially suffer from their location as well as insufficient institutional support.

Method and data
This article applies an exploratory case study approach [75].To address our empirical questions, we divided our empirical work into two steps: (1) definition of flood-exposed areas, and (2) semi-structured interviews and discussions with stakeholders in order to examine the factors for and against flood responses from the actors' point of view.
In order to understand the role of flood exposure for firm-level adaptation decisions, it was necessary to define urban areas with high levels of flood exposures.In a first step, we used information on severe flooding due to torrential rain at street level, which coincide with fluvial flooding from Storm Usagi in November 2018.Among the impacted districts, we opted for multiple case study areas consisting of six administrative districts (Fig. 1) to depict the implementation of flood adaptation strategies more holistically.Binh Chanh and Binh Tan provide peri-urban examples, while Tan Binh is an urban district mainly exposed to pluvial flooding.The eastern districts Nha Be (as a rural district in transition) and District 7 represent areas usually impacted by both fluvial and pluvial flooding and are thus integrated in urban flood prevention planning.District 2, east of the Saigon River, is likewise impacted, but not covered by recent flood prevention planning.
Information about flood depth and location for November 2018 flooding, derived from the Steering Center of the Urban Flood Control Program in Ho Chi Minh City, were combined with locations of manufacturing firms.On the basis of the 2016 Vietnam Enterprise Census (VEC) by the General Statistics Office of Vietnam [76], any manufacturing firm that has its production site within a "commune" (a sub-district administrative entity), which was either flooded by pluvial or fluvial flooding due to the heavy and prolonged November 2018 rainfall, is considered potentially exposed.In total, 5371 manufacturing firms are located within our case study districts.Almost one third (1, 432) of them is located in flooded communes and thus considered as potentially flood-exposed.Approximately 81,0% of exposed firms are small (less than 50 employees), 11,4% medium (50-200 employees), and 7,6% firms large (more than 200 employees).
In a second step, we conducted in-depth discussions with selected manufacturing firm representatives (see Fig. 1 for firm interview location), different stakeholder groups (government bodies, NGOs, consulting sector), and two business associations in HCMC in 2018.The aim of these explorative discussions and interviews was to gain basic local knowledge about firm exposure and adaptation strategies in HCMC.Based on this knowledge, and together with our Vietnamese colleagues, we developed a semi-structured questionnaire on the firms' economic situation, investment decisions, flood impacts, adaptation procedures, perceptions of future risk, and local risk management systems.The sample of firms for our interviews was randomly selected within the flood prone communes and included the full range of micro, small-, medium-sized, and large manufacturing firms.As we were interested in different perceptions and adaptation strategies in order to support the exploratory nature of the study and to identify possible differences across firm sizes, medium-and large-sized firms were sampled overproportionally.
We contacted manufacturing firms through lists of the Vietnam Chamber of Commerce and Industry (VCCI) and asked whether suitable respondents had experienced floods in the last five years.Our respondents were firm leaders, as strategic decisions in Vietnam, particularly in SMEs, are predominantly organized in a top-down manner [18].The interviews were mainly conducted with co-researchers in Vietnamese, translated in real time, and transcribed afterwards.The final results draw on 30 semi-structured interviews (10 micro/small, 11 medium, and 9 large firms, see Table 2) with manufacturing firms conducted in 2018 and 2019.
The qualitative material was analyzed in the following ways: first, to investigate firms' adaptive responses, we collected and examined information on implemented adaptation measures and those planned for the future.Based on theoretically derived considerations (see Section 2), we categorized these into action (proactive vs. reactive strategies) and inaction against flood risks.Second, we also grouped reported flood impacts according to direct, indirect, and wider effects [2], and mapped them against the type and intensity of response actions.Third, we then tested criteria for internal and external factors that influenced different manifestations of adaptive behavior (see Fig. 2).This criteria was inductively supplemented by additional emerging factors affecting decision-making.Here, we specifically paid attention to barriers to action and factors that could lead to adaptive inaction.

Flood impacts and adaptation strategies between action and inaction
More than half of the firms in the sample suffered from direct damage to buildings, machinery, transport vehicles, or even finished products, and almost all firms reported indirect and wider impacts in the last five years (see Table 3).The greatest reported impacts involved damaged buildings, machinery and materials, as well as the prolonged suspension of production, which ranged up to 30 days (SM-4) per year.Suspension of production was generally due to compound regular flooding events caused by storms and heavy rainfall during the monsoon season, which occasionally coincides with high tides.In our sample, many mediumsized firms are directly impacted by damage to buildings, machinery, transportation vehicles, raw materials, components, and products, and by a reduced or suspended production.This is mainly because they have permanently installed machines on the first floor, which can hardly be protected in the event of water ingress.Production is therefore quickly disrupted.Micro and small firms, for example manufacturers of mechanical products, are still more flexible in this respect.They usually work with very small machines that can be moved to areas of the premises that are not severly affected (SM-6, SM-7).
Two thirds of the firms interviewed were already implementing some sort of adaptation, while one third remained inactive.For those firms that were active, responses ranged from proactive and long-term ideas and measures to many merely coping actions.These latter allowed firms to become active only temporarly and reactively, and were often used in parallel.
Particularly SMEs facing indirect losses due to supply chain interruptions and postponed distribution of their products tend to react in a short-term manner (SM-1, SM-2, ME-4, ME-7).During the rainy season and high tide, it is difficult to preserve products in time for processing and export (ME-5).This situation leads to initially invisible but wider   impacts, such as the risk of non-compliance with contracts, the loss of customer trust, and possibly outright loss of customers [2,35].Private SMEs in the western districts Binh Chanh and Binh Tan, which produce light manufacturing products from paper (e.g., food packaging), food (processing), or textiles for the domestic market particularly suffered from business disruptions, damages, and losses, but failed to introduce proactive measures.One domestic firm producing plastic bags (SM-4), for instance, had elevated its production site once due to land subsidence in the neighborhood and regular heavy rain.However, the uplifted production site is still regularly flooded.The owner stated that the firm lacks the financial budget to implement additional adaptation measures and thus still suffers from the destruction of assets and stocks, threatening the firm's economic viability.Nevertheless, due to limited financial capacity, the firm is not developing further strategic adaptation plans.Another firm reported that their food packaging products are regularly damaged, and that they needed to stop production for a week in November 2018, which led to total damages of about 500-600 million VND (about 26,000 USD) (LA-2).Although the road in front of the company was raised, the impact of flooding has actually increased due to the lack of improvement in the drainage system.In response to flooding, the firm only blocks all doors and windows, removes wet products, and installs small temporary pumping systems.
On the opposite, we clearly see that export-oriented, foreign firmsoften located in industrial parks in District 7 and Nha Beare already much more protected from direct flooding impacts, but still implement long-term flood protection measures such as the installation of welldeveloped drainage systems to protect the factories from future flooding (LA-6, LA-4).With regard to indirect impacts on the delivery process, they respond by time-adjustments and, for example, by providing suppliers with water resistant plastic boxes to support the logistical processing of agricultural products during heavy rain (ME-8).
Contrary to our expectations, future-oriented and strategic adaptation measures were not intensively implemented or planned by floodexperienced firms that suffered from high exposure and impacts.Fig. 3 indicates, that most of the severely affected firms were more likely to cope in the event of flooding (e.g., SM-4, ME-1, ME-3, LA-2, LA-3) or invest into less capital-intensive strategies such as time-adjustments for employees or rearrangements within product delivery schedules (LA-7, LA-8, ME-6, ME-11).In contrast, firms that are currently the least directly impacted are already the best prepared for future flood risks.This result can be explained by two reasons: If adaptation is understood as a continuous process rather than a one-off action, firms that have already implemented adaptation measures are naturally, and in the long run, less directly impacted.Moreover, specific firm-internal characteristics and external conditions in the business environment (i.e., industrial park management, local flood management) influence firm decisions and opportunites to take adaptive measures in order to reduce their impacts.

Factors determining adaptive responses
Fig. 4 summarizes the internal firm characteristics and external conditions that determine adaptive responses.First, as assumed by McKnight and Linnenluecke [47]; Neise et al. [44]; and Wedawatta et al. [35]; firm size and financial liquidity play a predominant role for proactive adaptation.It is evident that smaller firms with less employees have difficulties in mustering financial resources to support effective responses.We find that the majority of SMEs with limited capital, and limited access to credits for long-term investments, usually cope directly after a flood event and hardly ever implement property-level flood protection.Once the flood occurs, firms continue their established business routines, placing products and other relevant materials in higher places to keep them dry, elevating machines and building barriers, such as with sandbags, in order to protect their production sites from water.They only react partially by modifying intended business plans (SM-8) through private hedging and small insurance (SM-2, SM-6).Although such organizational measures are generally seen as the basis for flood adaptation efforts [77], they do not protect production sites from direct impacts of flooding.Factory sites in Binh Chanh, Binh Tan, and District 7 are often situated at lower elevations than their adjoining streets, meaning that in the event of flooding water runs from streets into production facilities.As a result, firms with exposed and susceptible products such as flowers, paper products, or textiles especially express plans to raise their production sites, but they lack sufficient financial resources which could lead to surrender in the long run (SM-4, ME-1, LA-2).In this case, as Neise and Revilla Diez [22] argue on the basis of their research in Jakarta, that lack of proactive strategies runs the risk of constantly deferring actions to address flooding impacts to the future.
Second, the proactive behavior of micro-and smaller-sized firms strongly relies on the risk perceptions of decision-makers.Contrary to what Neise and Revilla Diez [22] have shown for Indonesian firms, the future economic impact of flooding in HCMC is commonly perceived by firm management, but this does not automatically lead to proactive measures (MI-2, SM-7).Thus, we find that some firm managers assess other business risks, such as shortages of workers, financial crises, and the emergence of competitors, as more relevant for their viability and competitiveness than flooding, especially in the case of flooding due to periodical high tides: "[…] we have more than 20 years of experience with floods and we have learned how to adapt.For us it is a kind of regular business, but we recognize that the magnitude of rain and tidal flooding is increasing" (ME-8).Whether there will be a future reconsideration with regard to flood risks,and the perceived necessity for protection among SLR, thus strongly depends on individual risk perceptions and the personality of the decision-makers.
Third, we find that external risk management systems strongly influence the extent of firms' individual flood adaptation.Our analysis shows that close contact with headquarters and subsidiaries overseas or strong relationships along the value chains drive the adaptive behavior of firms.Their engagement does not primary rely on actual flood impacts or local managers' decisions, but are much more influenced by guidelines and instructions.Firms enjoy the advantages of international exchange, knowledge spill-over regarding flood impacts, and floodrelated adaptation, such as the implementation of evacuation plans (LA-6).Moreover, they have well-trained staff capable to adopt, develop, and implement new techniques and strategies by participating Fig. 3. Flood impacts and firm responses.
R. Leitold et al. in training courses and workshops.Firms with international value chain linkages often possess sufficient financial resources to introduce proactive measures.A Japanese firm in the Tan Thuan industrial park, for example, reported that they lifted up their entire production site by 70 cm in order to prevent flooding inside company grounds.Additionally, they installed an internal drainage system proposed by their mother company and joined a regional business association in which adaptation possibilities are discussed with other firm managers (LA-4).In principle, locating production within an industrial park that has its own internal flood risk management is regarded as a panacea.Capital-strong companies in industrial parks tend to contribute more, and more often, to industrial park administration resources to protection against natural hazards and to support the expansion of local measures, such as drainage systems.Firms that are not in a position to efficiently prepare for future flood risks are often located outside industrial parks, but many seek to relocate to such parks to take advantage of infrastructural benefits (LA-8).Managers often mention that they already planned to relocate their production site, but hardly any interviewed firms had moved into less exposed areas: "I am not satisfied in the long term, because of not being in the industrial park and the flooding problems outside.This makes the company less appreciated by partners and ineligible for receiving foreign investments.For the future, we are trying to move to another area and upgrade the land" (LA-2).
Fourth, it becomes clear that the risk management undertaken by local governments and their public responsibility to reduce flood risks play a central role for firms' adaptive action.Thus far, firms usually view industial park administrations and public authorities as responsible for taking action, although they already appear overburdened to ensure flood protection and manage integrated flood adaptation strategies [78,79].If firms do not implement adaptive practices even if they are highly exposed to flooding, they often depend on public solutions and support mechanisms, which might be a risky strategy [80].Although the local government organizes some small conferences to raise public awareness about flood issues, information about major adaptation measures and practices do not enter into the business environments of micro-and small-sized enterprises.Most of the firms interviewed stressed that the state government does not take sufficient care of flood problems: "We think the government should be responsible, but they are just lifting up some streets in the city center or in other districts.They use our taxes for this kind of infrastructure" (ME-8).
Some urban development efforts by public authorities may even lead firms to no longer consider flooding risk and its reduction.Recently, there has been a trend of pushing for the relocation of labor-intensive production to areas outside the city center, which are often already highly flood-exposed [81].Thus, firm located in an official resettlement area may lead to inaction, especially in upgrading firm infrastructure and premises, and, ultimately lead to a wait-and-see strategy as a result (SM-4, ME-1).Moreover, some firms are aware of flood risk and are willing to adapt, but they have to wait for permission by industrial park managers and local district authorities to physically upgrade their production facilities (LA-2, ME-10).These bureaucratic procedures pose barriers to the implementation of sufficient flood protection.A precision-moulding firm, for instance, reported that they registered for an upgrading certificate, but had still not received permission after three years.Therefore, all flood adaptation efforts are just temporary solutions (LA-3).These firms even experience a double impact, as they are penalized by local authorities if oil or other hazardour substances leak out of their machines as a consequence of flooded production sites.These uncertainties about support and permission lead, if at all, to coping strategies in severe flooding situations, rather than the implementation of proactive strategies to reduce future risks more strategically.Beyond that, counterproductive and competing adaptation plans pursued by neighbors and local governments have been shown to lead to individual surrender.The most widespread challenge is the elevation of roads by the government.Firms in District 7 and Binh Chanh report that the effects of flooding have increased since the government raised some of the main roads in the city center (SM-3, SM-4, ME-1, LA-2).Consequently, some firm production sites became lower than the surface of the road, leading to major flooding that presented firms with new challenges that could not be solved by individual measures: "The problem of the floods is not solved by our own measures because the government is also lifting up the streets outside.Sometimes the flooding is so severe that our raised production facility is still flooded" (SM-4).
To meet the challenges of insufficient individual financial resources, capabilities, and support from public authorities, a collaborative adaptation approach [79] is gaining ground also in HCMC.However, the implementation of collective measures among neighbors, whether firms or other stakeholders such as NGOs or local authorities, has so far been an exception in our case study areas.One of the collective coping examples is the joint installation of sandbags and floodwalls of neighboring firms.Personal relationships and trust within local communities were identified as decisive factors for committing proactive collective action.One initiative responding to the lacking local infrastructure exemplifies such collaboration: "Previously there was an investor in our district who assured to invest in infrastructure, but this investment never happened.As a consequence, we decided to do it on our own.We lifted up parts of the street in front of our business together with our neighbors.Everyone who is located in this road payed for it.Some neighbors do not agree and refused to pay, so you see some parts of the street have not been improved" (MI-2).

Conclusion and outlook
Despite increasing interest in climate change, its economic impacts, and potential adaptation solutions, scientific understanding of whether and how firms take adaptation action is surprisingly thin.This contribution has reviewed relevant literature, developed a conceptual framework for sorting factors that might explain action and inaction, and provided empirical heuristics from firm-level adaptation decisions in one of the hotspots for climate change adaptation, HCMC in Vietnam.In particular, we were interested in understanding if adaptive action occurs more intensively among firms that experienced high direct impacts from flooding, and which internal firm characteristics and other external conditions may influence firm decisions to undertake flood adaptation measures.Our framework differentiated among proactive, reactive, and inactive adaptive behaviors.In order to explain these different behaviors, we looked into previous flood exposure and their impacts of firms, as well as into other influencing factors, ranging from internal dimensions such as the firm size and performance to external factors, such as support from public authorities and institutions.
Our analysis yields three main results: first, two thirds of the firms interviewed introduce proactive or reactive adaptation strategies while one third remained inactive.Most of the manufacturing firms cope only temporarily and reactively to prevent more severe impacts from flooding, but they do not pursue long-term strategies.Existing plans often focus on keeping people, products, material, and machines safe during flood events, rather than addressing systemic issues to avoid flooding damage more holistically.Second, experience with recurring fluvial flooding owed to high tides and high impacts to compounds from severe flooding do not automatically lead to strategic adaptation.Firms that are currently the least impacted are already the best prepared for future flood risks.Some of them have already implemented adaptation measures, meaning that they are already less directly impacted and have made provisions for future flooding.The firms currently most impacted are more likely to introduce mere coping strategies.High exposure to floods is often related to specific internal firm characteristics that act as barriers to forward-looking adaptation.Large firms with a stronger business performance and financial liquidity are generally more structured in their environmental responses as they tend to act proactively, even without direct experiences with severe flooding.In contrast, small firms with low financial reserves are often located in highly exposed locations, and they lack the capacity to adapt flood risks in the long term.This in turn increases their exposure and leads to compounding effects, which quickly endanger their future business viability.Moreover, the perception of managers plays a key role in adaptive decisionmaking as they have to assess the urgency of flood impacts compared with other business risks.Third, external risk management and support systems strongly influence the extent of firms' flood adaptation measures.The proactive engagement of well-prepared firms is mostly influenced by external requirements along the value chain which are facilitated by support and incentives from their headquarters and industrial park management.By contrast, those highly impacted microand small-sized firms are left behind, usually receiving no institutional support, and they are even exposed to competing adaptation plans by different actors.As a result, they mostly cope with severe flooding situations reactively, or even surrender, instead of implementing proactive strategies, despite their awareness of future risks.
Regarding our methodology, our findings show potential for future research, but also analytical limitations.The adaptation strategies discussed highlight local examples relevant to adaptation trends within the business environment of our case study area and cannot be automatically transferred to other field settings.Due to the time-consuming methdology, the number of interviews is limited.While our research presents an exploratory first step in the qualitative analysis of firm adaptation to floods, more in-depth research and data triangulation is needed to understand the dynamics and cross-relationships of both internal and external barriers and facilitating conditions.Moreover, further detailed analysis will be necessary to validate our findings and assess differentiation within adaptation strategies across different field settings, business landscapes, and sectors.
Looking into the future, our analysis shows that there is an urgent need for increased attention to firms when considering future adaptation to climate change, including firms' adaptive capacities as well as their possibilities and constraints for taking action.As the example of HCMC has underscored, SMEs are a particularly vital part of Vietnam's economy.Their long-term success is therefore of great societal relevance.Future risk governance thus needs to be directed to facilitate effective firm adaptation to climate change impacts.To achieve this, SMEs need to be aware of their flood risks and also be encouraged and supported in prioritizing risk management for their business activities.Local governments should provide information, effective guidance, and especially support mechanisms and incentives to finance further adaptation investments.To identify the right solutions, science needs to advance relevant knowledge, for example, in providing typologies for adapting firms to different contexts, examining the effectiveness of support tools, and analyzing the specific needs of SMEs, which typically do not receive the same attention as large international corporations.One particular focus should be on the question of how firms actor ought to actas agents within larger adaptation partnerships among state agencies, civil society, and private sector peers.Given the poor public adaptation capacities in many parts of the world, SMEs will have to bear a major part of their adaptation costs.But they will increasingly make their voice heard, pushing for integrated adaptation solutions and cost sharing with other parts of the society.
Garschagen, and Javier Revilla Diez was funded by the German Federal Ministry of Education and Research (BMBF) (grant numbers 01LZ1703D, 01LZ1703A).The contribution to this research by Van Tran was funded by the University of Economics and Law, Vietnam National University Ho Chi Minh City, Vietnam.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.Case study areas in Ho Chi Minh City (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.).

Table 1
Key understandings and definitions.

Table 2
List and coding of interviewed firms in Ho Chi Minh City.

Table 3 Direct
, indirect, and wider business impacts of flooding (percentage of firms in the sample).Direct impacts of flooding to flooded production sites & surroundings Damage to buildings,