Obtaining an insignificant variable which gets omitted

Using Variable1 from my dataset (survey), which represents how the respondent voted at the EU referendum 2016, I have created a variable "BREXIT", which I'm using as a dependent variable. BREXIT=1 (if respondent voted Leave at the EU referendum), BREXIT=0 (if respondent voted Remain at the EU referendum)

Using Variable2, representing which political party the respondent supports in the UK,  I have created a variable "UKIPsupport", which I'm using as one of the explanatory variables. UKIPsupport=1 (if respondent supports UK Independent Party), UKIPsupport=0 (otherwise).

My hypothesis is that those who support UKIP also voted 'Leave the EU' at the EU referendum in 2016 and I'm confident that this relationship should hold. However, "UKIPsupport" variable gets omitted when I'm running a probit regression (Stata says "note: UKIPsupport != 0 predicts success perfectly. UKIPsupport dropped and 121 obs not used"). Why can this be the case?

I'm attaching:

1) a table representing respondents who 'support UKIP' / 'don't support UKIP' versus respondents who voted 'Leave' / 'Remain'. There're no respondents in the sample who support UKIP and voted Remain, might this be the issue explaining why the variable gets omitted? Or is it something else?

2) probit regression results

Thank you