Evaluating The Road Accident Costs In Thailand Finance Essay
3.1 Framework of the Study
As stated earlier, the purposes of this thesis are to present the Willingness-To-Pay method to be applied to evaluate the road accident costs in Thailand. The methodological framework o achieve the objectives of this thesis is presented in Figure 3.1.
Hypothesis: WTP can be used to evaluate accident costs of Motorcycle Users
Using Willingness-To-Pay to evaluate cost
Risk of Motorcycle accidents
Contingent Valuation Method
Value of life of people using motorcycles
Relationship between Socioeconomic Characteristics and WTP
Relationship between Behavioral factor and WTP
Figure 3.1 The Methodological Framework of the Study
3.2 The Methodology of Accident Costs Analysis in Thailand
As stated in literature review, two methodologies have been widely used to evaluate the accident costs which including the Human Capital method and the Willingness-To-Pay method. In the past, Thailand has used the Human Capital method as a methodology for evaluating the accident costs. In this thesis, the researcher tried to propose the Willingness-To-Pay method as another methodology in order to evaluate value of accidents that occur on motorcycle users in Bangkok and surrounding areas.
3.3 Use of Contingent Valuation Method to Measure WTP
The important subjective elements that should be considered in evaluating the costs of human lives and injuries are “pain, grief and suffering, bereavement”. There have been many studies measuring a value of life, principally for use in benefit cost analysis. Value of life can be measured in several ways. In this study, the contingent valuation method is proposed as a method for measuring WTP.
Contingent Valuation (CV) is a commonly used method for measuring non market values. When CV is compared to revealed preference market valuation, the advantage of this method is that values can be explained in terms of underlying socio-economic conditions. Potential future events can also be evaluated, in contrast to the retrospective nature of market studies.
3.4 Design of Questionnaire Survey
3.4.1 Design of Contingent Valuation
A Contingent Valuation study of risk reduction should describe the nature of goods to be valued, and provide an evident description of risk and change of risk, and determine a realistic payment. The main attributes of survey followed the research conducted in the USA as follows.
Nature of goods: The goods in this research to be valued are fatal and non-fatal injury risks. Fatal and non-fatal injuries are described by an injury scale with three categories: slight, serious, fatal injury types. These categories are classified based on the definition defined by the Ministry of Public Health in Thailand.
Valuing questions are given in terms of private risk reduction. According to Strand (2002), the Willingness-To-Pay to reduce mortality risks may clearly involve the individual's valuation of others' death risk reduction, both family members and third persons.
Risk communication: Risk communication is one of the most important elements of Contingent Valuation studies about risk.
Payment vehicle and CV method: As a realistic payment vehicle, participants were asked about their willingness to pay for risk reduction.
The value of good or service in the Contingent Valuation technique is elicited through an elicitation method. The elicitation approach used in Contingent Valuation studies is different methods for asking willingness to pay questions: payment card (PC), dichotomous choice (DC) and open-ended (OE) approach. In this study, the payment card (PC) is adopted as method in the Willingness-To-Pay questions. The payment card contains a range of WTP values for the public good under question where individuals have to choose the maximum values that they have willing to pay.
3.4.2 Elicitation Method
Contingent Valuation elicitation questions can be designed as two basic forms namely: open-ended or closed-ended. For an open ended question, the respondent was asked to state the maximum amount at which he or she is willing to pay for the good that is being valued. For a closed-ended CV question, the respondent was asked whether he or she is willing to pay a specified amount presented as the value of the ameliorated service. For the closed-ended questions, the respondent is supposed to reply “yes” or “no.” Closed-ended questions have been the preferred form of elicitation question since it was introduced by Bishop and Heberlein (1979). In contrast, open-ended questions provide more information than closed-ended questions; and do not require econometric modeling to analyze, as the mean WTP values of respondents can be readily estimated by simple arithmetic. However, answering an open-ended question on a new commodity requires a higher level of cognitive demand on the part of respondents, because individuals are typically not accustomed to performing such tasks in daily life decision making.
3.4.3 Questionnaire Design
The questionnaire includes the questions about socio economics, household characteristics, and attitude towards the accident. In addition, the Willingness-To-Pay to increase the safety was asked by using the CV method. The example of questionnaires used in this study was shown in Appendix A.
The questions in the questionnaires can be classified in three broad categories as follows:
1. Valuation questions that are used to provide estimates of the relevant marginal rates of substitution of wealth for a small reduction in the probability of fatal and non fatal accidents.
2. Behavioral questions that are meant for measuring the degree of applied risk-taking through habitual practices on the road.
3. Personal questions that are used to solicit information on gender, age, income, and other details
The questionnaire includes the valuation questions in three scenarios. Each scenario was filled out as two elicitations which are the open-ended and closed-ended method. The objective of the survey was to evaluate the WTP of respondents in their own risk of death and injury. In stated of motorcyclist in Thailand, three scenarios were used to ask respondents, each scenario implying a private risk reduction. For the first scenario, the respondents were asked to imagine that he/she had to take a bus in order to go to Khonkaen province. The cost of the bus depends on condition of bus and driver in terms of safety. For example, it is assumed that for Bus A the cost of travelling to Khonkaen is 250 baht per trip, your chance of dying due to the road accident while travelling using this bus is 16/100,000 each year. Therefore, he/she was asked that “How much you willing to pay for preferable bus to Khonkaen to avoid mortality risk from 16 to 8 death persons every 100,000 persons”. For the second scenario, for example, the respondents were asked to imagine that he/she has to put the helmet. He/she has two choices for helmet type. The sample of the question is that for Helmet A, the cost of helmet is 400 baht per helmet, the chance to have severe injury due to the road accident is 26/100,000 each year. Therefore, he/she was asked that “How much are you willing to pay for better helmet quality to reduce the risk of severe injuries due to accident from 26 to 13 severe injury persons in every 100,000 persons”. For the third scenario, the respondents were asked to imagine that he/she has to choose the wheels to prevent the slippery of motorcycle. He/she has two choices for wheel types. The example of the question is that for Wheel A, the cost of this type of wheels is 600 baht , and the chance to have slight injury while accident occurred is 300/100,000 each year. Therefore, he/she was asked that “How much you willing to pay for better wheel quality to save risk of minor accident from 300 to 150 minor injury persons every 100,000 persons”.
3.5 Data Collection
Data collection is the most significant step since the data collected from the survey is needed to analyze for the Willingness-To-Pay and the value of life. In the data collection, it is categorized into two parts. The first part is the primary data which is collected through a field survey by using questionnaire. Secondly, the secondary data was obtained from the existing data such as the number of vehicle registration that is gathered from Land Transport Management Bureau, the number of people involved in motorcycle crashes and the number of deaths and injuries which is gathered from the Royal Thai Police and Ministry of Public Health. To obtain the data with good quality from the survey, the following criteria need to be applied (Aaker et al., 1995):
Population has been defined correctly
Sample size is a representative of the population
Interviewed respondents are available and willing to cooperate
Respondents understand the questions
Respondents are willing and able to respond
Interviewer correctly understands and records the responses
Respondents have required knowledge, opinion, attitudes, or facts.
However, in some situations, the answers from the respondents are not reliable due to the misunderstanding in the questionnaires. Therefore, to control the quality of data collection, the interview method was applied in this study. The respondents were asked and clearly explained by the well-trained interviewers for all questions included in the questionnaires to prevent the misunderstanding of the respondents.
3.6 Sample Size
Applying the Central Limit Theorem, a sample size can be determined from Equation 3.1 (Sudman, 1976). Assuming that coefficient of variation is 1.1, desired precision is 0.05, at 95 percent confidence level, the required sample size required is approximately 1859.
where n = Minimum sample size
Z = 1.96for a 95 percent confidence level
G = Plus or minus percentage points
c = = Coefficient of variation
Therefore, the sample size for this study is 2,400 individuals, separated in each province as 1,200 individuals for Bangkok area, and another 1,200 individuals for the surrounding areas.
The respondents are the people who live in Bangkok and perimeter and regularly use the motorcycles to travel. The cumulative number of motorcycle registered as of 31 December 2008 in the study areas are shown in Table 3.1. The selected locations for collecting the data are the offices of government agencies, private companies where most employees using the motorcycles, the schools or universities, the market or on the public streets . The interview was conducted in Bangkok, Nontaburi, Pathumtani, Samutprakarn, and Nakornpathom.
3.7 Description of Risk
The probability of risk of motorcycle users in Bangkok and perimeter areas can be classified as a decision tree in figure 3.2.
The value of probability at each section was calculated from the data obtained from various sources in Thailand as shown below:
Number of motorcycle registered: Land Transport Management Bureau
Number of people involve in motorcycle crashes: Royal Thai Police, Ministry of Public Health
Number of people injured or killed from crashes: Royal Thai Police
Number of people slight injury/severe injury/fatal: Ministry of Public Health
The probability of crashes, injuries, and property damage can be calculated based on the following equations
Figure 3.2 Event Tree for Bangkok Motorcycle Crashes and Injuries
Probability of Crashes =
Probability of No Crashes = 1-
Probability of Injury =
Probability of Property Damage Only = 1-
Probability of Slight Injury =
Probability of Severe Injury =
Probability of Fatal =
The risk proportion was analyzed from the probability of crashes of Motorcycle in Bangkok and perimeter areas. The methodology that was used to determine the risk for each type of accident is shown below.
The risk of accident = Pr. Crashes Pr. Injury Pr. slight or Pr. Severe or Pr. Fatal
Then, the risk reduction was calculated by taking the percentage of reduction multiplied by the risk of accident.
Table 3.1 The Number of Motorcycle in the Study Areas
Source: Statistics Sub-Division, Technical and Planning Group, Land Transport Management Bureau, Department of Land Transport
3.8 The Willingness-To-Pay Analysis
3.8.1 The Value of Statistical Life and Injury
Because people whose lives will be saved by a safety improvement cannot be identified in advance, the concept of statistical life has been developed (Kochi et al., 2001). VOSL (Value of Statistical Life) is usually expressed as willingness to pay for a change in risk divided by the change in risk.
VOSL = (3.2)
VOSL (or VOSI) is the total willingness to pay to avoid an expected occurrence of one fatality (or injury), if each person in the population of 1 million is willing to pay 35 baht to avoid a risk of 10-6. Since the total WTP also equals the mean WTP times the population size, the usual formula is obtained:
VOSL (or VOSI) = (3.3)
3.8.2 The Methodology to Analyze the Accident Cost
Mean of each elicitation question is completely difference in calculating. The calculation of mean can be analyzed by two methodologies as shown below.
18.104.22.168 The Open-Ended Question
The mean WTP values of respondents can be readily estimated by simple arithmetic. However, answering an open-ended question on a new commodity requires a higher level of cognitive demand on the part of respondents, because individuals are typically not accustomed to performing such tasks in daily life decision making.
22.214.171.124 The Closed End Question
Validity testing, which is vital to provide credibility to estimate WTP, still needs econometric modeling. The mean Willingness-To-Pay values of respondents should be estimated using econometric models (Hanemann 1984, Cameron and James 1987, Cameron 1988).
Therefore, the author used a step-by-step procedure to estimate mean WTP using a Probit regression model (Gunatilake, 2007). Mean WTP can be estimated for the entire sample. For Closed-Ended Elicitation question, it needs to use solution to determine mean WTP where is stated step by step as below.
Example to Estimate Mean WTP by Using Statistical Software
Step 1: Run a simple Probit regression in STATA software.
Step 2: To calculate the mean WTP, divide the coefficient on _CONS by the coefficient on dependent value and multiply by –1.
3.9 The Analysis of Influencing Factors Affecting the Willingness-To-Pay
This section summarizes the regression technique that has been used to analyze the influencing factors which affecting the Willingness-To-Pay of each individual. The influencing factors include various selected variables as shown in Table 3.2. In this thesis, the analysis was conducted in two parts. First, the individual Willingness-To-Pay (WTP) from a closed-ended questionnaire was analyzed by using the logistic regression techniques a statistical tool. The analysis was conducted to determine the relationship between independent variables such as the socioeconomic characteristics, the risk taking behaviors and the dependent variable which is the willingness of the motorcyclists in Bangkok and surrounding areas to pay for the value to survive from the road accident. In addition, the average values of the Willingness-To-Pay (WTP) of each motorcyclist from closed-ended elicitation were determined by the logistic regression as well. Second, an open-ended elicitation used the linear regression analysis technique to analyze for the aforementioned relationship. In this open-ended elicitation questions, the dependent variable or the value of the Willingness-To-Pay (WTP) is the continuous variable. The independent variables in this analysis are the same as the closed-ended analysis as shown in Table 3.2.
One purpose of this thesis is to use regression analysis to test effects of socio-economic variables on the Willingness-To-Pay.
3.10 Descriptive Statistics of Variable in Regression Model
According to answers of willingness to pay of respondents, they are given by closed-ended method which gives only lower bound and upper bound values in order to estimate expected value of willingness to pay method (E (WTP)), along with the use for considering factors that have effect on willingness to pay of individual. As for open-ended method, they are simply used analysis tool to analyze the willingness to pay of individual for acquiring whole mean willingness to pay value.
Variable names and descriptions are presented in table 3.2 which indicated abbreviation name to analyze in analysis tool.
Because quality variable is random sampling which has various values and discrete random variable, and continuous values, in statistical analysis need to use dummy variable.
Table 3.2 Descriptive Statistics of Variables in Regressions
Stated WTP in Thai baht, for a mortality risk reduction, severe and slight injury for open-ended ; for fatality, severe injury, slight injury
Age of respondent; Continuous variable
Binary variable coded as 1 if male and 0 if female
Binary variable coded as 1 if respondent has income less than 10,000 baht and zero otherwise
Binary variable coded as 1 if respondent has income between 10,001-20,000 baht and zero otherwise
Binary variable coded as 1 if respondent has income greater than 20,000 baht and zero otherwise
Binary variable coded as 1 if respondent’s family has income less than 20,000 baht and zero otherwise
Binary variable coded as 1 if respondent’s family has income between 20,001-40,000 baht and zero otherwise
Binary variable coded as 1 if respondent’s family has income greater than 40,000 and zero otherwise
Binary variable coded as 1 if respondent is government officer and zero otherwise
Binary variable coded as 1 if respondent is private employee and zero otherwise
Binary variable coded as 1 if respondent is student and zero otherwise
Number of respondent household; Continuous variable
Binary variable coded as 1 if respondent often rides a motorcycle and zero otherwise
Binary variable coded as 1 if respondent often wears a helmet and zero otherwise
Biary variable coded as 1 if respondent often rides using the wrong direction and zero otherwise
Binary variable coded as 1 if respondent often uses the speed >70 km/hr and zero otherwise
Binary variable coded as 1 if respondent often drinks alcohol during riding a motorcycle and zero otherwise
Binary variable coded as 1 if respondent is in the college level of education and zero otherwise
Binary variable, where 1 denotes that respondent has been ever involved in a road-traffic accident and zero otherwise
Binary variable , comparison between starting point of version 1 and 2 (1 if vesion2 and zero otherwise)
Binary variables coded as 1 for accepting to pay bid size X baht and zero otherwise ; for fatality
Binary variables coded as 1 for accepting to pay bid size X baht and zero otherwise ; for severe injury
Binary variables coded as 1 for accepting to pay bid size X baht and zero otherwise ; for slight injury
Binary variables coded as 1 for people who want to pay to reduce fatality risk and zero people who do not want to pay to reduce fatality riskTable 3.2 Descriptive Statistics of Variables in Regressions (Continued)
Binary variables coded as 1 for people who want to pay to reduce severe injury risk and zero people who do not want to pay to reduce severe injury risk
Binary variables coded as 1 for people who want to pay to reduce slight injury risk and zero people who do not want to pay to reduce slight injury risk
Difference between value of amount that respondents are willing to pay and starting point in each observation
Need help with your literature review?
Our qualified researchers are here to help. Click on the button below to find out more:
In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: