Description of variables
WTP is assumed to be influenced by the age, socioeconomic and health related characteristics of the respondents. Descriptions of variables under the study of WTP for green insurance were listed in Table 6.1.
Variables in the model are as follows:
Dependent variables = WTP for green insurance
The explanatory variables are classified into four categories:
The first category consists of the individualspecific characteristics, such as age of respondent, gender, education and marital status.
Table 6.1 Descriptions of variables for the study of WTP for green insurance
Variables 
Description of the variables 
Dependent variables 

WTP (binary) 
Are you willing to pay a premium for green insurance on a monthly basis? 
1 for a positive responsive and 0 for a nonresponse 

WTPA (quantitative variables) 
What amount are you willing to pay (WTP amount in rupees)? 
Independent variables 

Quantitative variables 

AGE 
Age of the respondent (in years) 
NOFM 
Number of family members 
FY 
Family income 
NOEMF 
Number of earning members in the family 
BID 
Bids given for regular and timely premium that the respondent is willing to pay for green insurance on monthly basis to finance healthcare cost 
Binary variable 

GEN 
Gender 
MARRIED 
Marital status 
FT 
Family type 
OWN H 
Own house 
EPYA 
Is there environmental pollution in your area? 
ICGB 
Do you have an interest in conservation of the green belt? 
DUHI 
Do you have health insurance? 
DISYA 
Does disease exist in your area? 
HOSPITAL 
Hospital type 
Categorical variables 

EDU 
Educational level 
PHKEC 
Do the people in your locality help in keeping the environment clean? 
DISFREQ 
What is the frequency of disease occurring in your family? 
DISEFMS 
Is any family member suffering from a disease? 
The second category includes familyspecific characteristics, such as family type and number of members in the family.
The ‘third category’ includes economic factors, such as, ownership of house, monthly income of the respondent’s family, number of earning members and insurance awareness.
The forth category includes pollution, environment problem and diseases, types of hospital services used and interest in conservation of green belt.
Explanatory variables were defined as:
X, = Age of the respondent in years X_{2} = Gender (Female=0, Male= I)
^ = Educational qualification of the husband (Illiterate = 0, Primary
^{3} school= I, High school =2, Graduate=3, More than graduate=4)
X_{4} = Marital status (Unmarried =0, Married = I)
X_{5} = Family type (Joint family=0, Nuclear family= I)
X_{6} = Number of family members
X_{7} = Own house (Own house =0, Rental house=l)
X_{8} = Family income
X, = Number of earning members in the family X, = Environment pollution in your area (No=0,Yes= I)
X_{0} = Existence of disease in your area (No=0,Yes=l)
У = Frequency of diseases occurring (Frequently= I, Moderately=2,
^{11} Rarely = 3, Not at alI=4)
X2 = Hospital type (Private hospitals =0, Government hospitals= I)
Xз = Interest in conservation of green belt (No=0,Yes=l)
Statistical analysis
The statistical analysis for the survey study was done in two parts: The first part include conducting the binary logistic regression analysis where the responses were ‘Yes’ or ‘No’ for WTP for health insurance and the variables influencing the decision (Bandara et al., 2013). In the second part, respondents who refused to accept the offered BID for green insurance were asked to state the maximum amount in rupees that they were willing to pay to obtain for health insurance, by using an openended question. This indicated a continuous measure of WTP for the goods or services, taking into account the monthly WTP as the dependent variable and finding the direction and magnitude of the determinant by using the multiple regression analysis.
Therefore, the logistic regression and linear regression techniques were applied in the study to test the WTP for green insurance. Linear regression and a general form of logistic regression were used to study the influence of the set of predicator variables on the WTP for green insurance.
Statistical analysis I
I: Binary logistic regression
Logistic regression is a common method to estimate WTP. This method was chosen because the response variable was in binary form and the predictor variable was a mix of continuous variables, binary variables and categorical variables. Since some of the respondents were not willing to pay the offered bid amount and their WTP becomes zero, which means that they did not want green insurance and had rejected the scenario. Therefore, the study aimed to identify the determinants of decisionmaking, that is, whether to pay for green insurance or not and what determinants helped respondents in decisionmaking, which was an insight research question. This part describes the method (Logit model) used to identify the determinants of participating in the green insurance scheme (Bandara et al., 2013). It provided an overview of the WTP, focusing on the CVM.
The specified model is given below.
The empirical model, measuring the probability of WTP for green insurance by the people of Delhi, is as follows:
P; is the probability of function and WTP_{j} is the WTP for green insurance, where 1 indicates that the individual is willing to pay for green insurance and 0 indicates otherwise. Xj is the vector of the observed characteristic of demand, which includes socioeconomic, attitudinal and behavioural variables; В is the vector with the corresponding estimated variable coefficient and the error vector E, consists of the unobserved random variable.
Binary logistic model:
where, p is the probability of WTP for green insurance (Yes = 1, No = 0).