Willingness of residents to invest in flood mitigation measures and to purchase flood insurance

Approximately 12,000 people live in the flood risk areas in Finland. New flood risk management plans were prepared for 21 areas as a part of the EU Floods Directive in 2015. In connection with this work, we conducted a survey in three flood risk areas in 2014. The total number of recipients was 4,710 and the response rate 39%. The study provided an indication of how local citizens perceive flood risk, and specifically whether they have already invested or would be willing to invest in private flood mitigation measures or to purchase flood insurance. We also examined how previous personal experiences of flooding affected their responses. The questionnaires were georeferenced in GIS and compared with flood hazard maps on a detailed scale. Over 27% of the respondents had experienced a recent flooding event, but only 9% had implemented flood mitigation measures and 10% knew that flood coverage was included in their insurance. A young age and having property at risk of flooding increased the willingness of respondents to consider implementing flood protection measures and purchasing flood insurance.


Introduction
In recent years, flood management has increasingly shifted to more integrated approaches in Europe (the EU Floods Directive; 2007/60/EC).The development of flood management in Finland also requires private households to take more responsibility for flood mitigation measures, i.e. through the use of flood protection devices and/or purchasing of ³IORRG´ insurance.
Some studies [e.g.1,2] have suggested that massive flood protection constructions are not sustainable solutions in the long term.Instead, greater emphasis should be placed on flood mitigation measures implemented by private households.This makes it even more important to recognize the factors that are significant in promoting such activities and to deliver useful information to residents.
A growing number of empirical studies have recently investigated the factors that drive private mitigation behavior, among which flood risk perceptions have been the most dominant [3].A large number of empirical studies have investigated risk perceptions and personal actions, but the relationship between perceptions and the behavioral response relative to preparedness is still unclear and controversial [4].Although a number of studies have addressed the issue of how to create incentives for households to control and reduce flood damage though insurance, empirical analyses of the effectiveness of such incentives are rare [1].Current empirical research on mitigation behavior lacks information on studies that deter people from adopting precautionary measures; the perceived effectiveness of flood mitigation measures has received relatively little attention in the current literature on mitigation behavior [1].

0DQ\ &DQDGLDQV IDOVHO\ EHOLHYH WKDW WKHLU KRPH LQVXUDQFH FRYHUV GDPDJH FDXVHG E\ RYHUODQG IORRGLQJ 2XODKHQ [6] DSSOLHG LQGHSHQGHQW VDPSOHV WHVWV WR LGHQWLI\ WKH GHWHUPLQDQWV RI UHVLGHQWLDO YXOQHUDELOLW\ WR IORRG KD]DUGV
The aim of this study was to assess the willingness of individuals to implement or consider implementing mitigation measures and to purchase or consider purchasing flood insurance in three significant flood risk areas in Finland.We were interested in revealing whether the same or different factors influence these two intentions to reduce flood risk.flood risk management plan of the area is a bypass channel in the Säpilä area.The channel has been part of the flood-related discussion for decades, and has therefore partially helped to maintain flood awareness among residents.
Kalajoki is a mid-sized (4,247 km 2 ) lowland river with a total length of 130 km and a mean discharge of about 30 m 3 /s.The Kalajoki river basin is mainly a rural area with a total population of 48,000, including the significant flood risk area of Alavieska±Ylivieska. Due to the low proportion of lakes (1.8%), the river basin is characterized by large intra-seasonal variations in discharge, resulting in frequent floods and low discharge in dry periods.Drainage and dredging projects in the river basin have accelerated the runoff.In addition, flood embankments have reduced the size of natural flood plains.The highest discharges occur in spring, but increased winter discharges may, for example, increase the risk of frazil ice dams (Veijalainen et al 2009).
The Lapväärtinjoki River catchment is rather small (1,100 km 2 , length of 75 km) but with a 160 m drop height and only 0.2% areal coverage with lakes.The average flow is 13 m 3 /s.The city of Lapväärtti is located on a lowland area close to the river delta, which makes it vulnerable to rapid floods, especially ice jam floods.Lapväärtti was not designated as a significant flood risk area, but after suffering from serious flooding in autumn 2012 and again from ice jam flooding in spring 2013, flood risk maps and management plans have been prepared for the area similarly to the other significant risk areas in Finland.New embankments were constructed and dredging was carried out.

Methods
The survey sample was drawn from a census register provided by the Finnish Population Register Centre.The questionnaires were sent to all the households in the significant flood risk areas of Kalajoki and Huittinen.The recipient was chosen to be the eldest person of the household and between the age of 18 and 75 years.The study area of Huittinen was extended so that all the households within 500 m from the flood hazard area (floods occurring statistically once in a thousand years) were also included.Because Lapväärtti is not yet designated as a significant flood risk area, the study area was the river basin and all of its households.
The questionnaire procedure followed Dillman ¶s protocol [7], with the recipients being contacted three times.The sample included a total of 4,790 households c , and 1,858 responses were returned (38.8% response rate).
The questionnaires included questions about the UHVSRQGHQWV ¶ H[SHULHQFHV Rf floods, opinions about flood protection measures, flood-related information and socioeconomics.There was also a description of the change in the compensation party from the government to insurance companies and a reminder that without insurance, the homeowner is now even responsible for severe flood damage.
The main interest was in determining whether the respondents had implemented any private flood mitigation measures and if they knew or assumed they already had flood coverage in their existing insurance, or if they might consider purchasing such insurance.These latter valuation questions were constructed in the contingent valuation method style, which is the most frequently applied method in valuing environmental assets [8].However, the monetary amounts of the willingness to pay (WTP) for private mitigation measures are not the focus of this paper. 7KH

Sample characteristics and descriptive statistics
The average age of all the respondents was 58 years (AGE, see Table 1).The gender distribution of the respondents might be skewed, as 39% of the respondents were women (FEMA).The questionnaires were sent to the oldest person in household About 17% of the respondents categorized themselves as forest or agriculture entrepreneurs (ENTR) and about 11% lived in an apartment house (APAR).
Based on the results, 41% of the respondents perceived themselves as living in an area with a high risk of fluvial flooding (RISK).According to GIS analysis, about one respondent out of ten actually lives on real estate at risk of inundation in the case of a fluvial flood (WRIS).The flood risk areas are delineated by local environmental authorities, who rely on hydraulic modeling and flood mapping, excluding sewage and pluvial flooding.This expert opinion may for various reasons differ from the view of the residents living in the area.Furthermore, it is likely that those with experience of flood damage or disturbance, or with a feeling of being at risk, were more keen to participate in the study.
More than one in four respondents had experienced damage or disturbance caused by flooding (DAMA).However, only 9% of the respondents had implemented flood protection measures (IMPL).This suggests that experience of damage or disturbance due to flooding may also be associated with minor harm, e.g.road closures or damage, rather than only the possibility of having water rising to WKH OHYHO RI RQH ¶V building.Approximately 8% of the respondents would consider implementing flood protection measures on their own real estate (WTP).Less than one-fifth of the respondents would consider extending their home insurance to include flood damage compensation (CONS), while only one-tenth of the respondents were aware that they already had such insurance (INSU).Over half of the respondents owned some property in danger of flooding (PROP).
One out of five respondents felt that their household could influence the damage caused by fluvial floods (ATTI).Almost half trusted in the ability of the authorities to plan flood mitigation measures (TRUS), and about 30% received new information about floods from the questionnaire (INFO).also more likely to have insurance policies with flood coverage, and vice versa.That is to say, these measures were complementary to each other, not exclusive.
There were also more variables explaining willingness to pay for property-specific flood mitigation measures than there were variables explaining the interest in flood insurance.For example, belief in the household ¶s capability to influence flood damage and receiving new flood-related information were significant in explaining the willingness to implement flood mitigation measures, but not to purchase flood insurance.In any case, this appears to be an incentive to inform residents living in flood-prone areas about how to protect their property against floods.The same notion has also been presented in other studies [1,2].However, landowners were more interested in insurance.
Furthermore, younger people were more interested in both mitigation measures and flood insurance policies.This can be seen as a promising sign and should be taken into account when planning information campaigns.Many respondents also wished for more information on how to protect their property, i.e. concrete measures and advice.Trust in the ability of the authorities to plan flood mitigation measures was not a significant factor influencing the willingness to either implement mitigation measures or purchase insurance.
However, as mentioned earlier, the average damage caused by floods in Finland is not particularly significant on a global scale.This is probably why there are currently no specific flood insurance policies, but water damage caused by flooding is included in most home insurances.However, if more severe flood incidents occur in the future, this might change the situation and create a new line of business for Finnish insurers.
Oulahen [6] reported increasing media coverage in Canada concerning the frustration among home owners on learning that their insurance policy does not cover damage caused by flooding.It remains to be seen whether something like this will also happen in Finland after the first major flooding and when people realize that the compensation practices have at least in some parts been tightened.This might more probably be the scenario if real estates suffering from the floods are uninsured and if building permission has been granted for areas that are too low-lying.In such a case, is the responsible party the owner or the administrator of the building permissions?

Figure 1 .
Figure 1.Location of the three study areas and examples of their flood risk characteristics.Whole flood maps for the areas can be found from www.environment.fi/floodmaps.

2 Flood management in Finland
b Each insurance company has priced its own products on the basis of risk assessment.In the future, actualized damage will affect premium prices.

3 The application 3.1 Three study areas
The variable DIST describes WKH GLVWDQFH RI WKH UHVSRQGHQW ¶V KRXVH from the nearest river or stream, i.e. the potential origin of fluvial floods.Most of the respondents lived near a watercourse, being approximately 445 m from the nearest river.We were interested in exploring how Finnish people have prepared for floods, whether it is through private mitigation measures or taking out insurance coverage in case of flooding, and what factors influenced their decisions.First of all, we found that people who had carried out flood mitigation measures or planned to do so were