I'm Alan Alder, and this is clear and vivid conversations about connecting and communicating.

There are two ways to think about the 3.5% rule. The first is that it's still true that a smaller than expected

number is probably a sufficient threshold to cross for active participation, really having a

powerful political, economic, and social impacts. It's not necessarily about getting a whole majority of the country

to rise up or something. It's about amassing a sufficient number of people engaged in active

participation to begin imposing real pressure on their opponents. It's within reach, basically,

of any movement to build that level of local participation. That's Erica Chenoweth, who with a colleague organized an international

team of scholars in identifying all the major efforts in the 20th century to change governments

by both violent and nonviolent protests.

Not only were nonviolent protests more effective than armed resistance,

but a surprisingly small percentage of the population, around 3.5% was enough in many cases to change the government. I think it's very interesting that before

you began to seriously study this, you thought nonviolent demonstrations, nonviolent actions

were well-meaning but naive. What flipped you over? What caused you to want to study it seriously?

Yeah, I'd say that's right. I thought they were noble ideas, but probably naive in the sense that

I could understand the impulse to hope that nonviolent resistance could be as effective or more

effective, but was not entirely convinced from the state of the empirical evidence at the time.

But I also had a sense that the empirical question hadn't really been fully examined to the extent that was possible,

which is to say, you know,

take a really large sample of cases that were comparable across history, some of which were

primarily using nonviolent resistance and the others that were using all of the different

armed revolutions with which everyone's quite familiar and actually look at them

side by side and find out which ones have the better track record and that actually hadn't been done

...