The reasons for the United States' crime drop in the 1990s have been the subject of considerable debate. As noted here, tough sentencing policies are one major reason, with even the anti-punishment side's experts grudgingly conceding that tough sentencing caused more than a quarter of the drop. Other estimates are higher.
One of the most controversial* hypotheses is that of Donohue and Levitt that the legalization of abortion by the Supreme Court's decision in Roe v. Wade, 410 U.S. 113 (1973) was actually a major factor. See Donohue & Levitt (2001) The Impact of Legalized Abortion on Crime, Quarterly Journal of Economics, 2, 379-420.
A new study of English data challenges that hypothesis.
This post on Vox, a European economics and policy blog, is by Leo Kahane from Cal. State, David Paton from Nottingham, and Rob Simmons from Lancaster. They plot out crime data for the U.S. and for England & Wales** in years since legalization, normalized to the starting point. A drop in crime rates where the first post-legalization cohort reaches its peak crime years appears in the U.S. data but not the English data.
No response has been posted yet on the Freakonomics Blog, but we have probably not seen the end of debate on the question.
* "All the death threats I got came from the left and not from the right." -- Steven Levitt.
** In the U.K., crime data for Scotland and Northern Ireland are tabulated separately. England and Wales together are one unit.