Rent assistance

[HughParsonage]

• "Compatibility" with other payments, i.e. I believe there is a different rate if the person is a recipient of different payments.
• Ensure up-to-date.
• Modelling capability (probably with respect to the SIH)

[MatthewKatzen]

I see that you've made the rent assistance an annual payment if you have an fy.year input but fortnightly if you have a Date input. I can see the logic behind having it fortnightly for Date as the rate may change if you convert that to an annual payment, however will this not cause some confusion? Would it be better to just have the fy.year a fortnightly payment too? Perhaps also adding a per parameter.

[HughParsonage]

per parameter is definitely needed. Agree to make it consistent. However, emit a message if per is missing.

[MatthewKatzen]

Eligibility criteria for rent assistance should be handled within parent payment function http://guides.dss.gov.au/guide-social-security-law/3/8/1/10 Most notably the criteria for young single payment receivers http://guides.dss.gov.au/guide-social-security-law/3/8/1/104

[MatthewKatzen]

Decided to make note in rent parameter regarding the board and lodging payments http://guides.dss.gov.au/guide-social-security-law/3/8/1/70 Also decided to add a logical argument to RA for the sharer's provision http://guides.dss.gov.au/guide-social-security-law/3/8/1/110

[HughParsonage]

In terms of the modelling capability, very broadly we want a function

model_rent_assistance : survey --> transformed survey

i.e. the function takes a survey data.table together with new parameters (like max_rate) to create a new version of the survey on which we can calculate basic statistics. Before we worry about the exact survey object, see if you can create a model_rent_assistance function using dummy data, like:

library(data.table)
library(magrittr)
library(grattan)

CJ(rent = 1:500,
     n_dependants = 0:3,
   has_partner = 0:1 > 0,
   is_homeowner = 0:1 > 0,
   lives_in_sharehouse = 0:1 > 0)

[MatthewKatzen]

Do you think model_rent_assistance should take params fy.year and date? It kind of goes against the point of modelling your own values but it will make comparisons between the model and the official rates easier.

[HughParsonage]

There needs to be a baseline, like baseline_fy in model_income_tax, so yes baseline_fy and baseline_Date.

[MatthewKatzen]

Note that currently the functions takes rates at a constant amount, however rates are recalculated every quarter or so

[MatthewKatzen]

Note that Prop_rent_paid_by_RA, Max_rate, and Min_rent are constant across all demographics which is not realistic. Perhaps have a dt input of all modelling params