Baxter & King as an endogenous variable

Thanks for noticing it! I tried now, but still get “ugly” IRFS: