Numeracy and politics

A weekend editorial complains that Trumpism is driving people away from religion. And, yes, churches that support Trump show something quite ugly about their own moral sense. But. What data we have shows that the nonreligious were growing long before Trump entered the political scene. Looking at the graph right, there is no obvious inflection around that timeframe. Which suggests the causes are broader than one man or even a political movement.

Though I suspect the author was well-intentioned, they make one of the most common mistakes in writing about trends. Before crediting a causal story based on an alleged change, always look at a graph of actual data, showing large enough timeframe to see what went on before.

Political pundits routinely distort data about what is happening. Many of Trump’s cheerleaders touted how he turned the economy around. The fact is that the US economy in his first three years continued on much the same course as under Obama. His pundits badmouth the current economy. The truth is that if Trump were now president and the economy were exactly the same, they would tout it as a huge success.

I suspect many people get fooled by such things, because they lack the technical experience to think about numerical data. It’s hard to make that excuse for Elon Musk, who is spreading bullshit about Texas and other states registering non-citizens to vote.

Update: And now I read about the turmoil team Trump is causing at the RNC:

The loss of talent may be particularly notable on the RNC’s data team – increasingly important in presidential elections – which is being relocated out of RNC headquarters in Washington and to the Trump campaign’s headquarters in Palm Beach, Florida. … Some staffers who declined return offers suggested they disliked the notion of living at work, which might also mean living alongside the most hardcore Trump campaign staffers.

People good at collecting, scrounging, and interpreting data generally aren’t ideologues.