,
7 tweets,
2 min read
Read on Twitter
Still a work in progress, but the monthly salience of racism/anti-black discrimination in the NYT does correlate fairly well (r=0.789) with perceptions of discrimination among white Democrats. Wish I could add more data points (there are 33 in the chart below), but these... 
...are what I could find thus far
Thinking of calling CBS and asking them to share with me some the data I was unable to access via the Roper database
Because some datapoints in the above are derived from few respondents (e.g. survey runs from april-may, with most of respondents interviewed in april) I converted everything to quarterly means. Result is a somewhat stronger relationship (r=0.83)
1992m4 was an outlier in the original graph, which it shouldn't have been (as nyt mentions of discrimination/racism-related terms spiked during this period due to the LA riots)
Quarterly also might make more sense theoretically, as respondent perceptions are likely to be similarly affected by media salience in the months that precede the interview dates
Added another datapoint (from 1998), which boosts the r to 0.842. Will keep updating this graph as I find more data
