Experimental Design Details
We conduct an online survey experiment with a sample of representative Australian households (aged between 25 years and 65 years and earn over AUD $52,000/year) who now have mortgages, have taken mortgages in the past, or plan to take mortgages in the near future. We aim for a 50:50 gender split with “prefer not to say” randomly assigned between males and females for quota purposes.
Subjects are randomly assigned to two treatment groups. For treatment 1, the primary outcome is the subject’s comfort level with a given loan size. For treatment 2, the primary outcome is the equivalence mapping between lump sums and monthly repayments of a given loan size.
In the first treatment group, we ask the subjects to consider taking a new residential mortgage loan of different sizes. The subjects are told to assume that the mortgage in question is the only residential mortgage they have, it must be fully repaid after 25 years, and that a 20% deposit has already been paid. The subjects are asked to indicate how comfortable they would be with the different loan sizes on the seven-point Likert scale. Two rounds of questions are asked. The mortgage loan is presented as a lump sum in one round and presented as a monthly repayment in another round (the order of the two rounds is randomised). For each round, the question is repeated over ten different loan sizes (Lump sums: $200,000; $300,000; $400,000; $500,000; $600,000; $800,000; $1,000,000; $1,500,000; $2,000,000; $3,000,000. Monthly repayments: $940; $1410; $1880; $2350; $2820; $3760; $4700; $7040; $9390; $14,090). The order of the loan sizes is randomised between monotonically increasing and monotonically decreasing.
In the second treatment group, we ask the subjects to consider taking a new residential mortgage loan of different sizes. The subjects are told to assume that the mortgage in question is the only residential mortgage they have, it must be fully repaid after 25 years, and that a 20% deposit has already been paid. Two rounds of questions are asked. In one round, the subjects are shown a mortgage loan as a lump sum and asked to use a slider to select the monthly repayment level of a 25-year loan that makes them feel equally comfortable/uncomfortable given current market rates. The question is repeated over five different loan sizes. In another round, the subjects are shown a mortgage loan as monthly payments and asked to use a slider to select the amount of a 25-year loan as a lump sum that makes them feel equally comfortable/uncomfortable given current market rates. The question is repeated over five different loan sizes. The order of the two rounds is randomised. We also randomly assign each subject one of the two sets of loan sizes as below, and within each set, the order of loan sizes is randomised between monotonically increasing and monotonically decreasing.
1. Lump sums ($200,000; $400,000; $600,000; $1,000,000; $2,000,000) and monthly repayments ($1410; $2350; $3760; $7040; $14,090)
2. Lump sums ($300,000; $500,000; $800,000; $1,500,000; $3,000,000) and monthly repayments ($940; $1880; $2820; $4700; $9390)