Willingness to pay for flood risk mitigation among residents living near river’s confluence in Yogyakarta, Indonesia

Both property buyers and owners are concerned about the impact of flooding on residential property price. Although several studies have investigated the possibility of property devaluation in flood-prone areas, there are still few that have looked into the effects of flooding on residential properties in Sriharjo, Imogiri, Indonesia. This study aims to examined housing characteristics such as structure, neighborhood, and environment on residential property value. We also created an interaction variable between flood and structural attributes to see whether the impact of flood inundation on property value varies depending on the age of the building, the size of the building area, and the number of bedrooms. Using the hedonic pricing model, we found that flooding reduces the market value of residential property value. The characteristics of property like land square, building square, building age, and number of rooms have influence on property price. The characteristics of neighborhoods are also influence to property price such as distance to city, distance to main road, distance to school and distance to the river. Only one environmental characteristic which influence to property price is height of flood inundation. We discovered that, on average, a rise in the amount of flood inundation decreased property prices in the research region by.015% by using a two-stage estimating approach to evaluate these associations. We calculated IDR 698,500 as the marginal implicit cost of reducing flood inundation. According to subsequent estimates, homeowners would be ready to pay an extra 0.685 percent for a decrease in flood inundation. The study’s findings will help property owners better understand the elements that contribute to property depreciation due to floods. This paper also suggests flood insurance plans since flooding is a significant issue for real estate owners.


Introduction
The long-term viability of development initiatives may be impacted by natural disasters.The devastation of economic resources, such as the capital stock, infrastructure, natural resources, and people, has a direct effect.The government's budget may be affected by natural disasters.According to Hoffman and Bukroff (2006) [1], when a calamity strikes, the government must immediately raise funds for the reconstruction fund and the running budget.The consequences of natural disasters have been the subject of several research.The Global Facility for Disaster Risk Reduction report (2012) [2] demonstrates how flooding is a natural calamity that harms society, the environment, and the economy.One of the most common natural catastrophes that significantly damages people, livestock, and possessions is flooding, according to reports.Floods have the potential to destroy homes, buildings, and other assets.People were forced out of their houses, upsetting their usual lifestyles.In accordance with the Queensland 1314 (2024) 012061 IOP Publishing doi:10.1088/1755-1315/1314/1/012061 2 Floods Science, Engineering and Technology Panel (2012), depending on how serious the flood is or how mild it is, different effects of flooding on properties may occur.Flood recovery relies on how long it takes and how much it costs in countries that have suffered significant flooding.Property values may decrease as a result of this.Minor floods, however, frequently need for quick damage repair [4,5].
The use of a hedonic pricing technique to complete evaluation has been controversial ever since Ridker and Hennings' study in 1967.Finding functional forms that ultimately result in a fundamental functional form is the central problem.Numerous studies have employed hedonic property pricing methodologies to measure the connection between flood hazards and residential property prices [6][7][8].These studies frequently reach the conclusion that, on average, homes in floodplains sell for less than identical homes outside of the floodplain.
Numerous rivers pass through the Yogyakarta region, including the Winongo, Code, and Gajah Wong rivers.The latter is a tributary of the Opak river and empties into the south coast at Depok Beach in Bantul Yogyakarta after meeting the Oyo river, which originates in the Gunung Kidul region.Due to their strategic location and potential for disaster, highly populated urban centers are especially susceptible to the effects of climate change.Imogiri District, particularly in the Sriharjo Village, is one of the Yogyakarta region's areas that experiences flooding relatively frequently.
According to DIBI BNBP (2017) [9], there were 979 flood occurrences in Indonesia in 2017 out of a total of 2,862 incidents.399 episodes of flooding occurred in DIY in 2017.According to Pusdalops BPBD DIY (2017), there were several flood occurrences in the Special Region of Yogyakarta between 2015 and 2018, with the majority taking place in Bantul Regency (99 incidents), Sleman Regency (57 incidents), Kulonprogo Regency (35 incidents), Yogyakarta City Regency (33 incidents), and Gunung Kidul Regency (14).According to Pusdalops BPBD Bantul, 2017 in Iwan 2019, there were seven flood disasters in Bantul Regency in 2017.These disasters caused damage to flooded buildings, casualties, washed-away houses, submerged rice fields, damaged roads, destroyed bridges, and other submerged infrastructure.
Data from BPDB DIY (accessed March 18, 2019) indicates that flooding has affected 27 sub-districts in Yogyakarta.The flooding caused 5,046 inhabitants to be uprooted and dispersed in various locations.Kulon Progo Regency is one of the regions devastated by the flooding.Bantul Regency is the next worstaffected region, followed by Bantul.Villages in Imogiri's sub-districts such Wukirsari, Sri Harjo, Selopamioro, Girirejo, and Imogiri were impacted by the floods.The Celeng River overflowed, causing this flood.landslides as well as floods.Five persons perished in this landslide tragedy and were buried under the debris, according to statistics from BPBD.The eastern side of the landslide covered the region below, specifically the Kedung Buweng hamlet, along with the graves of the Mataram rulers (Sunny and Setyowati 2020).The following is the overall scenario in Sriharjo Village, Imogiri District, Bantul Regency: 1. Sriharjo Village; 2. Sriharjo Village covers around 502.36 Ha; 3.There are about 9000 people living in Sriharjo Village; 4. The primary source of income for those who live in Sriharjo Village is farming.Conditions generally in Sriharjo Village, which is part of DIY Province's Imogiri District, Bantul Regency.There are 13 hamlets in the village of Sriharjo, including the hamlets of Miri, Jati, Mojohuro, Pelemadu, Sungapan, Gondosuli, Trukan, Dogongan, Kentos, Ngrancah, Pengkol, Sompok, and Wunut.Wunut is the largest region, and Dogongan is the smallest.Males outnumber females in the population in practically every age range.The population in the productive age group is larger than the population in the non-productive age group.
According to its geographic position, Kalurahan Sriharjo is bordered by the confluence of two of the area's major rivers: the Opak river, which rises from Mount Merapi, and the Oya river, which rises from the Gunung Kidul region.This region is extremely susceptible to floods because of its location so near to the river.Additionally, the Celeng River, an Oya River tributary, being present has contributed to this position.There is no monetary value to matters like flood management, river beautification, and water quality.These are referred to as non-market products in economics.Municipal planners and politicians who want to vote on legislation based on the interests of their constituents have access to critical information from analyses of public interest in and willingness to pay for flood management and ecological restoration..Both revealed preference and expressed preference approaches can be used to determine WTP.The former, such as the travel-cost and hedonic price methods, examine the purchase of related goods in the private market place to determine demand for goods or services.The latter, such as the contingent valuation method, examines the purchase of related goods in the public market place to determine demand for goods or services, and choice experiment techniques examines the individual's stated preference for goods or services in comparison to alternative goods.The approach chosen depends on the study's objectives, the availability of the data, and the specific economic values (use and/or nonuse values) that must be considered.For non-market commodities like recreational fishing, the revealed preference methodology is still often employed, but the stated choice method is used for non-market goods like predicting discount rates in developing nations [10].Samaraweera et al., (2010) [11] contend that the uncertainty in economic decision-making and the financial and physical hazards associated with flood losses result in a significant economic cost.Knowing the precise cost of earlier catastrophes is crucial in this situation.According to the Royal Institution of Chartered Surveyors (RICS) (2010) [12], among other factors, the development of more modern low-lying structures and changes in weather patterns are to blame for the increase in flood occurrences.floods have harmed several homes that were not previously at risk of floods.The assessment of flood damage has been the subject of various research [13][14][15].This study shows that there is widespread agreement that residences inside the flood plan are less valuable than homes outside the floodplain or in a flood-prone location.
According to the line of reasoning put forward by Zhou et al. (2013), the area's sensitivity and hazard susceptibility to the investigated adaption technique are determined by a flood risk assessment framework, which is used to conduct the flood risk study.They emphasize that in environmental economic analysis, hedonic valuation is employed to account for at least the majority of the externality values associated with metropolitan water infrastructure.
After providing this justification, Scawthorn et al., (2006) [16] contend that in order to estimate building damage from flood damage, the damage module needs two inputs: the occupancy type of the building and the elevation of the first floor, which typically includes the design level; and flood depth.
These elements are required for this research since its focus is on the impact of flooding rather than calculating the overall damage caused by flood hazards.However, if the cost of flood damage to the property is determined by its real cash worth or replacement cost value.The study employs a hedonic model to assess the value of changes in observable variables (proxies) that are highly linked with the underlying hazards under consideration because the risk of natural catastrophes is not immediately observed.The price of replacing a building's current component is known as the value of replacement cost (RCV).In order to calculate a hedonic land pricing model, Okagawa and Hibiki (2011) estimated da phases within the Tokyo metropolitan government.The study's findings show that areas without a danger of floods have much higher land prices.This analysis of earlier studies has revealed a widespread consensus that severe flooding harms property.
An analysis of costs and benefits may then be performed using the projected value of a natural resource.The findings of a study on how flooding affects property prices in flood-prone locations, namely in Indonesia, are still unclear.This study investigates the variables affecting real estate costs in Sriharjo, Bantul Regency, Indonesia.We conducted a survey to learn more about the home's, property, area, and surroundings.We investigated the marginal willingness to pay for reducing the flood risk in the 426 homes in Bantul Regency, which is prone to flooding.

Study site
This research is in Sriharjo Imogiri Bantul, Indonesia.The selection of research locations was based on the vulnerable area of flood because near river's confluence.According to BPDB DIY data, there are 27 sub-districts in Yogyakarta affected by flooding.Around 5,046 residents were displaced due to the floods and scattered in several points.

Survey and administration
The sampling technique used strategic random sampling.We found 436 respondents to be surveyed.The questionnaire is divided into four parts: ▪ The first part is the structural characteristics (land square, building square, number rooms, building age, ownership), ▪ The second part is the neighborhood characteristics (distance to river, school, road, health facility, city, tree) ▪ The third part is environmental characteristics (flood inundation, rain frequency) ▪ The last part is property price.

Data Analysis
The hedonic pricing approach calculates the affects of environmental features on property value in order to estimate the welfare effects of environmental assets and services.The implicit demand function for the environmental goods may be estimated by tracking changes in property values, according to the hedonic pricing theory, which makes this assumption that if environmental quality varies, property prices will similarly change.Accordingly, hedonic pricing are described as the implicit prices of the attributes and are made known to economic actors by the observable prices of differentiated items and the precise quantities of characteristics attached to them [17].
Data collection on all factors that affect the sales prices of the property is required in order to estimate a hedonic pricing function.The explained variable, home price, is viewed as a function of neighborhood, environmental, and structural factors; these data sets are related to residential regions.
The Hedonic Model's estimation process was split into two phases [18].The hedonic property price function was estimated in the first step, and implicit prices were calculated for each observation.For certain sets of environmental factors, the hedonic pricing function was used to create the implicit demand function, also known as the marginal willingness to pay function.The steps are described below.

Specification of the Hedonic Price
Function.The hedonic pricing function, which is computed using a straightforward least squares regression model, connects the price of residential property to the property's structural, neighborhood, and environmental attributes [19].The hedonic pricing function is calculated as follows by using this generic specification and removing unimportant variables to make it more precise.ln PPRICE = a0 +a1 LANDSQ + a2 BUILDSQ + a3 ROOM + a4 BUILDAGE + a5 OWNERSHIP + a6 DISRIVER + a7 DISSCHOOL + a8 DISROAD +a9 DISHEALTHFAC + a10 TREE + a11 FLOODINUND + e where, ln PPRICE is natural log of property price, LANDSQ is land square, BUILDSQ is building square, ROOM is number of rooms, BUILDAGE is age of building, OWNERSHIP is ownership of building, DISRIVER is distance to river, DISSCHOOL is distance to school, DISROAD is distance to main road, DISHEALTHFAC is distance to health facility, TREE is existence of tree in neighborhood, FLOODINUND flood inundation, and e is error term.This function's implicit marginal price is determined by the partial derivative with regard to air quality.This cost is the extra sum that the household, assuming all other factors remain the same, would be prepared to spend in order to select a home with lower levels of air pollution.The following is an estimation of the marginal implicit price: 2.3.2.Specification of the Implicit Demand Function.The estimated implicit pricing for various locations are in accordance with each person's willingness to pay (WTP) for a marginal unit of an environmental good.The level of a feature is chosen by the person when their Marginal Willingness to Pay (MWTP) for that characteristic equals its implicit marginal price.Then, the implicit pricing as a function of flood inundation, other socioeconomic traits of people, and a demand shift variable, such income and family size, are regressed to produce the inverse demand function.The implicit pricing function taken into consideration for the final calculation is after the inconsequential variables have been eliminated using the approach of trial and error.where IMPPRICE is implicit price; INCOME is income of household head; and FAMSIZE is number of family member.

Results and Discussion
Surveying a total of 436 people, it was discovered that the average residence costs 184.16  Among the 'neighborhood variables', six major onesdistance from river, distance from school, distance from main road, distance from health facilities, distance from the city, the existence of tree in neighborhoodare also considered for this analysis.The distance between the household's home and the closest connected location is what was calculated.River proximity is 602 meters on average.The average distance to the closest school is 600 meters.796 meters separate you from the main street on average.
The income of the household head and the number of family members are two examples of "socioeconomic variables."The average household has four family members and earns 2.089 million rupiahs annually.The average flood inundation in the hedonic pricing model used in this study is 8.83 millimeters (Table 2).In order to estimate the hedonic pricing equation, we use the assumption that there is a negative correlation between the price of real estate and the environmental feature of flooding.The model's structural characteristics, such as the amount of land and the size of the buildings, are all anticipated to have positive relationships with the price of real estate.Property prices are negatively connected to neighborhood attributes like the distance to the hospital and school.Typically, it is believed that as a property's distance from a hospital or school grows, its price would drop, however as a property's distance from a city grows, its price will also grow.Following the application of these hypotheses to the model previously mentioned, the parameters are estimated using the ordinary least squares approach and are shown in Table 3.
The variables land area, building, number of rooms, distance to river, distance to main road, distance to health facilities, and distance to city have a significant impact on the selling price of the property at the level of 5%, according to Table 3 which has undergone partial t-testing.The price of a property is significantly impacted by the presence of trees and the height of the flood inundation by 1%.At a level of 10%, the factor of property ownership has a considerable impact on the selling price of the asset.According to this interpretation, the land area variable has a big impact on how much a home sells for.The variable area of land has a variable coefficient value that is positive, the larger the land area, the selling price will increase.This research is in line with what [20] said that land area is very influential in determining the selling price of a house.The wider the land size of a property, the property has a high price.The large size of the land means that the property has a large number of rooms that can attract individuals to buy the house.The building area variable has a significance value and the coefficient value that is positive.In other words, the larger the building area of the house, the selling price will increase.According to statistics, Sriharjo Village homes' selling prices are positively and significantly impacted by the building area.This research is in line with that conducted by Ismail et al (2019) [21] which states that the building area affects the selling price of the house, the wider the size of the house, the higher the price.The large size of building area means that the house has a large number of rooms that can attract individuals to buy the house.Other supporting research was carried out by Egbenta (2015) [22] and Rabassa and Zoloa (2016) who said that the building area was very influential in determining the selling price of the house and resulted in a high selling price.The regression model's variable age of the building results in a negative correlation between the variable age and the house's selling price.This implies that the price of a property will decrease the older the structure.The variable age of the building in this study has a statistically significant effect with a negative coefficient on the selling price of houses in Sriharjo Village.The results of this study are in line with Ismail, et al (2019) and Lee (2015) [23] explaining that the longer the age of the house building, the selling value of the house will decrease.This is because old or old houses require more maintenance and repair costs, as well as maintenance and repairs on building structures or electrical problems.In addition, houses with old building age have architectural forms that are less desirable according to current conditions so that they can reduce their selling value.According to study by Rabassa and Zoloa (2016), the same notion holds true that a building's age reflects its quality and condition; the better the state of a structure, the greater the selling price of a home will be.
The variable number of rooms affects the selling price of the house with the positive relationship.The price of the property will rise in direct proportion to the number of rooms owned.The selling price of homes in Sriharjo Village is significantly and statistically positively impacted by the variable number of rooms in this study.This study's variable room count has a statistically significant and favorable impact on the selling price of homes in Sriharjo Village.The results of this study are in line with research conducted by Egbenta, et al (2015) and Rabassa and Zoloa (2016), the large number of rooms contained in the house has an effect on the increase in the selling value of the house.This is because the large number of rooms in a house will provide a lot of space for each family member and the room is an important room in the house.This research is also supported by Rahmawati (2017) [24] that consumer considerations in buying a house are seen from the number of rooms available because the average consumer who buys a house is a consumer who has married so that the number of rooms is adjusted to family members.Therefore, the more the number of rooms, the higher the price for the house.
The variable distance to the city center produces a significance < 0.01 (1%) which describes H0 is rejected and H1 is accepted, meaning that the distance to the city center affects the selling price of the house.The variable has a negative relationship between the distance to the city center and the selling price of the house.The city center is a place where all transportation access to the area and to the city is easily reached and the location contains the center of government, trade and offices.Therefore the location of the house that is closer to the city center is worth more than the location of the house that is far from the city center.The variable distance to the city center in this study statistically resulted in a negative and significant relationship with the selling price of houses in Sriharjo Village.This negative relationship means that if the location of the house is far from the city center, the selling value of the house will decrease.This is in line with research by Rabassa and Zoloa (2016) that the location of the distance from the house to the city center is an important factor where it is easy to reach all cases in the city center because it is the center of all activities such as government centers, exhibitions, entertainment and offices.This statement is also supported by Votsis and Perrels (2016) [25] who say that increasing the distance from the location of the house which is far from the city center results in a significant decrease in the selling price of the house.
The variable distance to the main road has an effect on the selling price of the property with negative sign of 0.038.The main road is a facility that makes it easier for the community to access all the surrounding facilities.Therefore, the location of the house that is close to the main road has a higher value than the location that is far from the main road.The variable distance to the main road in this study has a significant effect with a negative coefficient on house prices in Sriharjo Village.A negative relationship means that if the location of the house and the distance to the main road are far, the selling value of the house will decrease.This is supported by research conducted by Meldrum (2016) [26] and Jung &Yoon (2016) [27] which have a negative effect.A house that is close to the main road will certainly have the added value of very easy transportation access.However, the proximity of the main road to residential homes also has several weaknesses, namely pollution and noise from these weaknesses, making the value of the house down.This is contrary to research by Rahmawati (2017) consumers are more likely to consider that polluted air can disturb and cause discomfort.
The variable to school has a significant effect on the selling price of the house.The variable distance to school in this study has a significant effect on the selling price of houses in Sriharjo Village.This is not in line with the research of Saptutyningsih and Basuki (2012) which states that being far near the school has no effect on the selling price of the house.This discrepancy in influence is due to the fact that education centers are now evenly distributed in every kelurahan and the numbers are quite adequate.As is the case in Sriharjo Village, which has 11 public and private schools that include Elementary Schools (SD) to Vocational High Schools (SMK).Because the distribution of schools is evenly distributed, it makes it easy for everyone.According to Lee (2015)'s research, which supports the findings of this study, the selling price of a home was significantly impacted by the distance to the local educational facility.The sale price of the property is impacted by the property's changing distance from the river.This is in line with the research of Saptutyningsih and Basuki (2012) [28] which states that being far near the school has no effect on the selling price of the house.This discrepancy in influence is due to the fact that education centers are now evenly distributed in every kelurahan and the numbers are quite adequate.As is the case in Sriharjo Village, which has 11 public and private schools that include Elementary Schools (SD) to Vocational High Schools (SMK).Because the distribution of schools is evenly distributed, it makes it easy for everyone.The results of another study that contradict this research were conducted by Lee (2015) who said that the distance to the education center had a significant effect on the selling price of the house.
The flood inundation height variable in this study has a significant influence on the selling price of houses in Sriharjo Village.This study is in line with the results of research by Egbenta, et al (2015) who researched in Lokoja, Nigeria which stated that houses located in flooded areas experienced a reduction in the selling price of the house on average.There is a study from Lee (2015) which states that flood inundation has no significant effect on the selling price of houses in South Korea.This happens because when there is a flood, the height of the flood inundation that overflows into the house area does not overflow into the house, the duration of the inundation does not take so long to recede so it does not really affect the damage to the house.In addition, the frequency of flood events in Sriharjo Village only occurs if there is a long period of heavy rain in the upstream area of the river.According to Zulkarnain et al., (2018) [29], the damage caused by flooding is characterized by a flood inundation height of more than one meter which can damage building structures and the duration of the flood which takes up to 12 hours.Therefore, the height of the flood inundation has no effect on the selling price of houses in Sriharjo Village.The implicit marginal price function for environmental products may be deduced from the first derivative of the hedonic pricing function.Table 4 provides descriptive statistics of implicit marginal pricing for 436 observations.Therefore, IDR 698,500 is determined to be the marginal implicit cost of reducing the flood risk.This finding makes it abundantly evident that, in addition to structural and neighborhood factors, flood inundation has a significant role in influencing the demand for real estate transactions in Imogiri Bantul, Indonesia.
As was previously indicated, the implicit marginal price is regressed on the quantity of environmental products purchased as well as other socioeconomic characteristics, such as the people' income level, to estimate the inverse demand curve in its second stage.The main conclusions are that the implicit marginal price function's first derivative with regard to flood inundation is negative (-.685), indicating declining implicit marginal pricing for rising environmental quality.This indicates that in the research region, a 1 centimeter decrease in flood risk results in a.685 percent gain in implicit property price.All other factors are significant at the 5% level of significance, with the exception of land square, building age and ownership, distance to school, main road, and healthcare services.It's noteworthy to notice that for every additional meter of distance from IOP Publishing doi:10.1088/1755-1315/1314/1/01206111 a river or city, the implicit property price increases by 082% and.1542%, respectively.The implicit property price is strongly correlated with building square footage and the number of rooms.
We have now covered the estimate of the hedonic pricing function and how it affects the price of real estate.The findings make it abundantly obvious that homeowners are prepared to pay for reducing the danger of flooding.Therefore, it is vital to calculate the welfare advantages they would receive from buying property with a lower flood risk.

Conclusions
The risk of flooding that occurred in the Sriharjo Imogiri caused the selling price of property in the area to be influenced by several risk factors, because of its location as a flood-prone area.Based on calculations using the hedonic price method approach, it can be concluded that the property characteristics of land area, building, building age, number of rooms affect the selling price of the property.Neighborhood characteristics such as distance to the city center, distance to the main road, distance to schools, distance to the river, also affect the selling price of property.Likewise, environmental characteristics, especially high flood inundation, greatly affect the selling price of property in Sriharjo Imogiri.
We discovered that, on average, property prices decreased as flood inundation levels increased by using a two-stage estimating approach to evaluate these correlations.The estimations also showed that families would be ready to pay for less flooding.In brief, the investigation showed that households in Sriharjo Imogiri, Bantul, responded favorably -which prone to flood -between environmental quality and property prices.Therefore, the local government together with the community need to review the spatial layout of settlements on the banks of the river, especially the three major rivers in Yogyakarta.This is intended so that the value of property in disaster-prone areas, one of which is a flood disaster, still has a high selling value.Public awareness in disaster-prone areas also needs to be increased in the context of mitigating disaster risk, both natural and man-made disasters.Local wisdom is needed to create resilience to disasters.

Table 1 .
The relation of variables and references + Ismail,dkk (2019), Rabassa dan Zoloa (2016), Egbenta,dkk (2015) Number of rooms (ROOM) + Rahmawati (2017), Rabassa (2016), Egbenta,dkk (2015)) Building Age (BUILDAGE) -Rabassa dan Zoloa (2016), Egbenta,dkk (2015), Lee (2015) + Ismail,dkk(2019), Egbenta,dkk(2015), Saptutyningsih (2011) Dependent variable: Property price (PPRICE) million rupiahs.One of the key factors influencing home pricing is neighborhood.A square meter of land is 346.5 square meters on average.Buildings typically have an area of 108.4 square meters.The amount of rooms a house has is another element that affects the price of residential real estate.The study area's homes typically have three rooms, on average.The building's age is varied, with the lowest number being 1 year and the highest value being 207 years.The average age of the building is 26.69 years.

Table 2 .
Descriptive statistics of the survey participants (n = 436)

Table 3 .
Regression result of Hedonic Price Function Table 5 contains the results.