Can someone explain to me why 189, 190, and 491 have quite different refusal rates?
Using the aggregate data from July 2024 to October 2025, I got the following:
For 189, the primary refusal rate was 3.14% if withdrawals included and 3.18% if withdrawals excluded in the calculation;
For 189, the overall refusal rate was 2.63% if withdrawals included and 2.66% if withdrawals excluded in the calculation;
For 190, the primary refusal rate was 1.48% if withdrawals included and 1.50% if withdrawals excluded in the calculation;
For 190, the overall refusal rate was 1.31% if withdrawals included and 1.32% if withdrawals excluded in the calculation;
For 491, the primary refusal rate was 2.36% if withdrawals included and 2.45% if withdrawals excluded in the calculation;
For 491, the overall refusal rate was 2.16% if withdrawals included and 2.25% if withdrawals excluded in the calculation.
The calculation formulas are (grant_p) / (grant_p + refusal_p + withdrawal_p), (grant_p) / (grant_p + refusal_p), (grant_o) / (grant_o + refusal_o + withdrawal_o), and (grant_o) / (grant_o + refusal_o + withdrawal_o) for the four scenarios, with "p" and "o" standing for primary and overall here.
With over 20K cases for 189, 40K cases for 190, and 30K cases for 491, I believe these numbers should be statistically significant.
Why are we seeing the lowest refusal rates for 190, medium refusal rates for 491, and highest refusal rates for 189 applications?
For all the subclasses, the refusal rates are 3% (ish) or lower, less than 1 out of 30 applications. No need to assume any concrete cause and effect OR patterns varying by subclass. I can float one explanation: for 190/491 states check claims first so it gives more assurance, where for 189 it goes straight to CO.