Health shock and preference instability: assessing health-state dependency of willingness-to-pay for corrective eyeglasses

Table 4 Willingness-to-Pay for Corrective Eyeglasses

	[1]	[2]	[3]	[4]	[5]
	Average WTP	Shock induced changes in WTP	Standard Error	95% Confidence Interval	Shock as a percentage of Average WTP
Entire Sample	664.6 ***	–	11.32	642.5–686.8	–
Unmatched Without-Health-Shock Group	596.8 ***	–	12.86	571.6–622.1	–
= 1 if diagnosed with refractive errors	–	257.8 ***	25.44	208.0–307.7	43.20%
1 to 1 Matched (Replaced) Sub-Sample	808.9 ***	–	17.41	774.7–843.0	–
Matched Without-Health-Shock Group	762.4 ***	–	26.78	709.9–814.9	–
= 1 if diagnosed with refractive errors	–	116.8 ***	32.93	52.3–181.4	15.32%
1 to 1 Matched (Not Replaced) Sub-Sample	773.9 ***	–	16.45	741.7–806.2	–
Matched Without-Health-Shock Group	691.9 ***	–	23.62	645.6–738.2	–
= 1 if diagnosed with refractive errors	–	171.3 ***	31.37	109.8–232.8	24.76%
Kernel Matched Sub-Sample	772.0 ***	–	16.49	739.6–804.3	–
Matched Without-Health-Shock Group	687.4 ***	–	23.70	640.9–733.8	–
= 1 if diagnosed with refractive errors	–	176.0 ***	31.43	114.4–237.6	25.60%
Radius Matched Sub-Sample	757.8 ***	–	16.15	726.1–789.4	–
Matched Without-Health-Shock Group	659.9 ***	–	22.76	615.2–704.5	–
= 1 if diagnosed with refractive errors	–	199.0 ***	31.02	138.2–259.8	30.16%
Mahalanobis Metric Matched Sub-Sample	771.4 ***	–	15.86	740.3–802.5	–
Matched Without-Health-Shock Group	690.0 ***	–	22.14	646.5–733.3	–
= 1 if diagnosed with refractive errors	–	165.1 ***	30.33	105.6–224.5	23.93%
Weighted by Propensity Scores	792.8 ***	–	45.82	758.9–826.7
= 1 if diagnosed with refractive errors	–	234.6 ***	31.52	172.9–296.4	29.59%

(a) Average WTP - is the mean predicted value obtained through constant-only model where the log-likelihood of the model is iteratively maximized given a mean predicted value, (b) Shock induced changes in WTP - is the marginal WTP attributable to the health shock (i.e. being diagnosed with refractive errors); it is the coefficient obtained through fitting interval regression model with a dummy variable which is equals to 1 for experiencing the health shock and 0 otherwise, (c) Each of the sub-samples (i.e. obtained using the respective matching techniques) contains 1116 matched respondents of With- and Without-Health-Shock groups, (d) The Matched Without-Health-Shock Group contains 558 respondents selected as a viable match for the 558 With-Health-Shock group respondents, (e) Shock as a percentage of Average WTP - is obtained by calculating the Shock induced changes in WTP as the percentage of the Average WTP of the entire/sub-samples; it describes by how much the average WTP for eyeglasses of the With-Health-Shock group is higher compared to that of the Without-Health-Shock group, (f) Asterisks indicate statistical significance (*** p < 0.01)

ISSN: 2191-1991