John Lott the Debunker

I've heard a lot of people criticize Freakanomics and its claim that abortion reduces crime because the children who would have grown up to be criminals were aborted 18 years ago. Most of the criticisms I've heard center around attacking utilitarian logic and that we shouldn't kill one fetus to save several murder victims.

John Lott's Freedomnomics has the first criticism I've seen that focuses on real statistical data. He begins with an interesting finding:
Roe v. Wade was decided in 1973.
For children born between 1969 and 1973, by the time they reached the ages of 18-25, the rate of murder was normal.
For children born between 1974 and 1978, by the time those children reached 18-25, their rate of murder was dramatically higher.

Lott's explanation is that legalizing abortion has led to more children being born out of wedlock, since it caused increased promiscuity among young people. The number of children born into single parent homes increased, and those children started committing lots of crimes.

I'm not sure how convincing this long logical chain is, but that's mainly because I don't know if abortion actually caused lots of children to be born out of wedlock. This sounds correct, but he's the first person I've encountered to make this claim.

Anyone have thoughts on whether or not there is a causal connection between making abortions more accessible and having more single parent families? The economic explanation presented is that abortions reduce the "cost" of sex since both parties no longer have to worry about raising a child, and that this has therefore led to an explosion in irresponsible sex and actually more teen pregnancy.