Talking Together

Just picked up a copy of Talking Together: Public Deliberation and Political Participation. Yes, note the titular colonicity -- gotta have a colon to qualify as serious academic work, otherwise it's just hackery.1

Anyway, this book focuses on deliberative democracy. Basically, you sit a bunch of people in a room, educated them in an unbiased fashion about the issues, then poll them to get results theoretically better than the kind of uninformed opinion you'd often get from a random survey of U.S. adults.2

I'm burying the lede. I've only skimmed the book so I can't give an overall judgment, but there are concepts that matter to me: political engagement, participation, attention, and of course my old favorite, knowledge (which plays a bit part, not a starring role). In this case, there is a set of multiple regression tables in Chapter 5 with participation as the dependent variable and a host of independent variables tossed into the recipe. These ingredients are grouped as demographic factors (race, gender, etc.), social capital (belong to organization, religious attendance, etc.), political capital (efficacy, trust, knowledge, attention), and the last in the model -- deliberation.

Now the lede: deliberation still explains unique variance even after controlling for every possible other variable in the social science universe.3 Now the secondary lede: knowledge sometimes retains its explanatory power, even after all these controls. In some tables yes, some tables no.4

So knowledge leads to participation? Couldn't it be the other way around?

There are some obvious causality issues here, and the authors rightly mention them, so I'm not going to get into that methodological mess. Just skimming, the book is sound, thoughtful, and a worthy addition to the participation/social capital/deliberation literature.

---
1 Like this blog
2 That these people differ from regular people is hardly surprising
3 Okay, not every variable. Beer consumption is oddly missing
4 Attention, however, offers zilch to the model