# Statistical analysis part II: Multiple regression analysis

The current study reveals that about 59.11 per cent of the respondents have accepted the BID amount offered to them randomly, showing their WTP for green insurance. The remaining 40.89 per cent of the respondents decline to pay the BID value offered to them. The latter group are asked in the form of an open-ended question to specify the maximum amount they are willing to pay for the proposed scheme. Thus, in the open-ended question, the respondents are asked to state the maximum amount of money they are willing to pay to get the benefit of green insurance. That would give a continuous measure of the amount that respondents are willing to pay for green insurance and reveal how the amount of WTP varies with the specified socio-economic and other related determinants. Hence, it would help in finding the direction and magnitude of the determinants by using multiple regression analysis. This study uses the Ordinary Least Squares regression to determine the relationship between the actual amount respondents are willing to pay for green insurance and their socio-economic determinants.

## Model specification

The empirical specification of the model is:

WTPA = Amount respondents are willing to pay

The multiple regression analysis (Table 6.11) shows that the maximum amount of money that the respondents are willing to pay is positively

Table 6.11 The amount respondents are willing to pay for green insurance, using the regression model

 Model Unstandardised coefficients Standardised coefficients T % 6 Std. Error Beta (Constant) -317.915 202.322 -1.571 .116 Age of respondent -59.065 23.076 -.078 -2.560 .011** Gender 138.746 41.135 .081 3.373 .001*** Educational level 152.718 20.302 .341 7.522 .000*** Marital status 79.427 45.208 .043 1.757 .079* Family type 18.942 64.166 .01 1 .295 .768 Number of family members -65.508 32.777 -.074 -1.999 .046** Own house -47.866 45.035 -.028 -1.063 .288 Family income 51.977 19.867 .080 2.616 009*** Number of earning members in the family 1 16.659 27.529 .105 4.238 .000*** Environment pollution in the area 202.539 77.783 .123 2.604 009*** Existence of disease in the area 1 13.912 67.562 .069 1.686 .092* Frequency of occurrence of disease 35.530 17.564 .047 2.023 .043** Hospital type -24.661 42.599 -.014 -.579 .563 Interest in conservation of green belt 7.356 32.868 .007 .224 .823

Dependent variable: Amount respondents are willing to pay (in Rupees) * indicates 10% level of significance.

• ** indicates 5% level of significance,
• *** indicates 1% level of significanceSource: Primary survey

Table 6.12 Test statistics of the amount respondents are willing to pay for green insurance, using the regression model

 R 0.733 R squared 0.S37 Adjusted R square 0.529 F 73.254 Significance of F .000 Degree of freedom 14

Source: Primary Survey

related to the level of education, family income, number of earning members in the family, existence of disease in their area, frequency of occurrence of disease, number of visits to private hospitals, interest in conservation of green belt and environment pollution in their area. Male members are more WTP for green insurance than female. At the same time it is also found that married people are more WTP than unmarried. In fact, the level of education, marital status, family income, number of earning members in the family and environmental pollution in the area are seen to be positive and highly significant factors in influencing the amount of money respondents are willing to pay for green health insurance.

Table 6.12 shows the test statistics. R is the multiple correlation coefficient that tells us how strongly the multiple independent variables are related to the dependent variable. The regression coefficient R value is 0.73, which shows that there is a high positive correlation between the dependent variable and all the independent variables in the model. R square gives the coefficient of determination. It implies that (0.537), that is, 54 per cent variation in the dependent variable (amount respondents are willing to pay for green insurance) is explained by variations in the independent variables in the model. The adjusted R square value is 0.529, which gives us the idea of how well the model generalises. Ideally, we would like its value to be the same or very close to the value of R square, which is 0.537 in the model. The test statistic from the results of the multiple regression model demonstrates that the model is quite robust with F statistic value of 73.254, with over 99 per cent level of confidence.

## Derivation of demand curve for the amount respondents are willing to pay for green insurance

Table 6.13 shows the amount of money respondents are willing to pay for health insurance in the open-ended question. The WTP for green insurance is higher at the lower premium amount of insurance than that of the higher premium amount. This shows that health insurances with lower premium are highly demanded by the respondents for health insurance.

Table 6.13 Amount respondents are willing to pay for green insurance, using the open-ended question

 Amount in Rupees Percentage of Respondents WTP 1-250 28 251-500 25.8 501-1,000 23.8 1,001-2,000 16.2 2,000-5,000 6.2 Total 100.0

Source: Primary survey

In Figure 6.3, rhe X-axis represents the percentage of respondents and the Y-axis shows the amount of money that respondents are willing to pay. The demand curve for the amount that respondents are willing to pay for green insurance in the open-ended question is a downward sloping curve with a steep slope. The demand curve is relatively elastic, which shows that respondentsâ€™ demand for green insurance is very sensitive to the change in the amount to be paid.

Figure 6.3 Demand curve for amount respondents are willing to pay for green insurance

Source: Primary survey