Financial decision-making behaviors of Ethnic Tibetan Households based on mental accounting

Table 13 Regression results from a ranked ordered logit model identifying the factors that influence the repayment order of different loans

Alternative (base: temple loan)	Loan from friends and relatives		Bank Loan		Government's anti-poor loan
Variables	Coeff.	Std.Err.	Coeff.	Std.Err.	Coeff.	Std.Err.
Credit village	0.372	0.505	0.150	0.603	0.321	0.595
Temple	0.013	0.348	− 0.008	0.338	0.291	0.371
Skill	− 0.405	0.536	− 0.454	0.422	− 0.462	0.437
Religion	− 0.004	0.166	0.387	0.321	0.589*	0.331
Income	0.025	0.043	− 0.050	0.050	0.085	0.066
Str income	− 0.811	0.649	− 0.812	0.777	− 1.512	0.972
Dowry	− 0.009	0.037	− 0.167**	0.070	− 0.198**	0.088
Age	0.020	0.015	0.018	0.018	0.005	0.017
Gender	− 0.120	0.440	− 0.201	0.460	0.302	0.322
Married	0.789**	0.393	0.874*	0.468	1.164***	0.401
Health	− 0.228	0.231	− 0.565*	0.333	− 0.397	0.316
Edu	0.143	0.145	− 0.034	0.193	0.150	0.196

Dependent variable is the repayment order rank of four kinds of loans (1 = temple loan, 2 = loan from friends and relatives, 3 = bank loan, 4 = government's anti-poor loan). It's worth noting that among the independent variables, str income means the proportion of agricultural income in total household income. Standard errors are estimated by village clustering
*, **, and *** represent significant at 10%, 5%, and 1% level, respectively