The correlation between Hispanic CD percentages and the swing to the GOP in CA CD's in 2022

[Torie]

Regarding the purported Hispanic swing to the GOP this cycle, I decided to do a deep dive into CA. Below is a link to a scattershot diagram with a trend line overlay, which compares the Hispanic percentage in each CA CD to the percentage that the GOP candidate (where there was a GOP candidate – the was none in 5 CD’s), ran ahead of the Trump 2020 percentage. There was a correlation (with a rather weak T stat).

My conclusion is that the correlation was more due to very low, indeed shockingly low, Hispanic turnout, rather than Hispanics switching their partisan preference vote. One weird thing is that the Dem percentage in the Hispanic portion of the San Fernando Valley had a huge collapse (a whopping 17.63% swing from 76% to Biden to 59% for the Dem CD candidate (both percentages rounded)), along with the typical turnout collapse. What is up with that? The lowest swing, 0.66%, was in, you guessed it, the very high turnout coast north of San Francisco.

[Torie]

First, thank you for explaining the CA-29 anomaly. That brings to 7 districts, elections where there was no GOP candidate on the ballot. Those 7 districts have been excluded from the analysis.

As leavened by candidate quality/incumbency, it appears that the GOP swing in CA from Trump 2020 to 2022 Congressional races was driven more by the disproportionate failure of Biden voters to turnout than a swing by Hispanics per se. While there is not a statistically significant correlation between the Hispanic percentage of a CD and GOP swing percentage, there most definitely is one between the percentage decline in turnout in a CD and GOP swing from 2020 to 2022.



Here is a scatter diagram chart. I have labeled some of the CD’s of interest, particularly those farthest from the trend line. It does appear that the bulk of the CD’s farthest from the trend line have to do with incumbent candidate quality.



Here is a chart with the raw data from which the above is derived. I have identified CD’s that don’t fit into the pattern (e.g., low turnout/low Hispanic percentage, high Hispanic percentage/ low swing, high turnout/high swing, low turnout/low swing).