About the first: Correlation does not imply causation. It's very difficult to be sure if the same communities that makes laws to force abstinence-only education also make other laws like forbidding than teenagers get sexual counseling in hospitals without their parents. (And it's imposible to be sure unless you kidnap all the children and put them in two isolated islands in randomized groups.)

Also, they classify the states according to the legislation about abstinence-only education:

0: no mention of abstinence

1: mention, but no promotion

2: promotion of abstinence

3: stressing abstinence as the fundamental teaching standard

It's very interesting that the pregnancy rates for 0 is very similar than 2. 1 is better, 3 is worse. Perhaps the states that don't mention it apply a strong implicit social pressure?

About the second one: I didn't expect this result.