Skip to main content


Table 10 Main effects MNL model estimates

From: Attribute development and level selection for a discrete choice experiment to elicit the preferences of health care providers for capitation payment mechanism in Kenya

 Preference estimatesWillingness to accept (WTA)
AttributesLevelsCoefficient (robust se)value (robust se)
Payment schedule1 month− 0.0895**
3 months
6 months
12 months
Timeliness of paymentsDelayed0.4808**
− 1580.4597**
Capitation rate per individual per year1200 shillings0.0003**
2400 shillings
3600 shillings
4800 shillings
Services to be paid by the capitation rateCapitation rate pays for consultation only.−0.0360
Capitation rate pays for consultation and drugs only
Capitation rate pays for consultation and lab tests only
Capitation rate pays for consultation, lab tests, and drugs
Performance requirementsHospital receives base/fixed capitation rate0.0540
− 177.6008
Hospital receives base/fixed capitation rate + bonus for improved performance (e.g. improved quality)
Opt-out −0.2319
Model fit statistics   
Log-likelihood at convergence− 250.6159  
Log-likelihood (final)− 222.5633  
Adjusted rho-squared at convergence0.09  
Akaike Information Criterion457.13  
Bayesian Information Criterion478.21  
number of decision makers (n)31  
  1. s.e. - Robust standard errors in parenthesis. Asterisks denote statistical significance at *** 0.1%, ** 1%, and *5% level