That is a shocking take in my opinion, one that borders on delusional. Statistics are the result of specific metrics collected by people who chose what specific data points to collect, the methods of collecting those metrics and chose the methods of presenting the data. They can reveal interesting aspects of reality that aren’t otherwise obvious and can depict a fairly accurate representation of reality as a whole if they are created in ernest using sound data collection techniques, but I’m pretty sure that the most qualified data scientists will disagree with the statement that “statistics are reality”. Especially if anyone in control of any part of that process has significant motivation for them to depict something specific.
Statistics are only meaningful when you put them into context of their intent, limitations and error rate.
And even if the statistics hold true in aggregate, it’s not the full picture and can’t accurately describe or predict individual experiences. Perception is anecdotal, so it is not a perfect depiction of reality either. But if perception does not match the data, it’s an indicator that the data might be suspect.