The data suggests no such thing. As the report underlying this article describes (https://static1.squarespace.com/static/5b6df958f8370af3217d4178/t/6011e68dec2c

The data suggests no such thing. As the report underlying this article describes (https://static1.squarespace.com/static/5b6df958f8370af3217d4178/t/6011e68dec2c7013d3caf3cb/1611785871154/NYU+False+Accusation+report_FINAL.pdf https://static1.squarespace.com/static/5b6df958f8370af3217d4...), there's no high-quality data available on question of whether conservative views are disproportionately censored. The data shows that conservative accounts don't have systematically suppressed engagement, but that's fully compatible with the idea that there might be some views those accounts know they aren't allowed to talk about. At the end the report calls for Twitter to be more transparent in their moderation decisions, precisely because the lack of transparency makes it hard to figure out answers to this kind of question.