,
18 tweets,
5 min read
Read on Twitter
1/n I've previously shared a bunch of data linking social/digital media use to the 'Great Awokening' (If you missed those threads, you can find them here ). Data I've been looking at from the general social survey points in the same direction.
2/n I begin by illustrating trends in white internet usage, which increased tremendously across the board since the GSS first included the measure. Whereas white liberals, on average, reported surfing the net for 5 hours a week in 2000, this figure jumped to 15+ hours in 2018.. 
3/n which is significantly (albeit not considerably) greater than all other ideological subgroups.
4/n Unsurprisingly, this trend coincides with marked increases--particularly so among liberals--in the percent of respondents listing the internet as their primary news source. Whereas 20% of white liberals listed the internet in 2000, almost 60% did in 2018
5/n So to what extent does this trend in internet usage map onto increases in woke responses to race-related survey questions? To examine this, I begin with an item asking respondents what they think is the main cause of black-white inequality. Those who selected..
6/n ..'discrimination' serve as the dependent variable, which I regressed onto 3-way internet use x ideology x year interaction + controls for party affiliation, age, sex, socio-economic status, and religious service attendance.
7/n What's interesting is that among white libs and mods, the average effects of internet use on 'discrim' responses turn significantly positive only over the past decade--a period which, as I've noted elsewhere, is marked by the social media boom + rise of hashtag activism.
8/n Tellingly, we also see that the predicted probability of a 'discrim.' response from those placing in the bottom quantile of weekly internet use remains more or less flat across time, while the odds of this response from the top quantile shows a significant secular increase.
9/n Zooming in now on 2018, whereas a white lib placing in the bottom quantile of net-use has a predicted 49% chance of giving a 'discrimination' response, this estimate climbs to 76% among those in the top. A similar but significantly smaller pattern also obtains among mods.
10/n Our next dependent variable comes from a question (taken from the standard 'racial resentment' battery) that reads as follows: 'Irish, Italians, Jewish and many other minorities overcame prejudice and worked their way up. Blacks should do the same without special favors.'
11/n Respondents who 'disagreed' with this statement were coded as 1, and all others 0. I then regressed this variable on the same model as the previous. As you can see, a similar pattern obtains: the average effects of net use on a 'disagree response' becomes sig. positive...
12/n ...only within the past decade, and only among white liberals and (if lesser so) moderates. A look at the predicted margins reveals a widening disparity between the white liberal responses in the bottom and top quantiles of weekly internet use.
13/n By 2018, a white liberal placing at the bottom quantile has an estimated 20.6% chance of giving a 'discrimination' response as compared to a 60.8% probability for those scoring at the top.
14/n Insofar as immigration is a 'racialized' political issue, we can expect to find a similar relationship between net-use and support for increasing immigration (which has skyrocketed among white liberals over the past few years ).
15/n And indeed we do.
16/n In 2018, a white liberal scoring in the bottom quantile of net-use has a predicted 25.4% probability of supporting increased immigration as compared to a 58% chance among those placing at the top.
17/n All told, these findings are consistent with those I've reported elsewhere, and lend strong evidence to the notion that white liberal digital media consumption (including 'outrage porn') is at least partly accountable for the White liberal 'awokening'. That said, I need to..
18/18 ...figure out an experimental design that can test the causal argument.
