Polarization is estimated separately for the treated question hours and for the other sessions. Thus, we can estimate the effects of broadcasting on speech polarization by comparing levels in these different groups before and after the start of broadcasting. (4/8)

We observe a positive effect on the speech differences between the government and the opposition, suggesting that broadcasting question hours can exacerbate language use. However, we do not observe that the broadcasts of question hours would affect left-right polarization. (5/8)