Abortion and Crime, Revisited (Ep. 384 Update)

Episode Summary

Executive Summary: The episode revisits the controversial Freakonomics thesis that legalized abortion helped drive the dramatic drop in U.S. crime by reducing the number of unwanted births. It explains the original evidence, backlash, methodological critiques, and a 20-year update showing the relationship remained strong. The discussion broadens into how complex social outcomes often have multiple causes and how public debates oversimplify causality.

Main Topics: Supreme Court leak and abortion policy context (Priority: 5/5): The episode opens with the leaked draft opinion suggesting Roe v. Wade could be overturned, framing abortion as a renewed national and political flashpoint and motivating the rebroadcast of the earlier Freakonomics investigation. Original abortion-crime hypothesis (Priority: 5/5): Steve Levitt and John Donahue describe how legalized abortion may have reduced later crime by lowering the number of unwanted children, who were statistically more likely to face adverse life outcomes and criminal involvement. Evidence, methodology, and critique (Priority: 5/5): The researchers explain their quasi-experimental approach using variation across states and birth cohorts, then address criticism over a coding/description error and clarify that the core finding held up after correction. Alternative explanations for falling crime (Priority: 4/5): The episode compares abortion to other factors in crime decline, especially lead exposure, and emphasizes that multiple causal forces can operate simultaneously rather than only one explanation being true. Public reaction and political/moral discomfort (Priority: 4/5): The paper triggered hostility from both pro-choice and pro-life audiences because it linked an emotionally charged private decision to a utilitarian public outcome, and also got entangled with race and eugenics concerns. Updated findings and policy implications (Priority: 5/5): A later analysis using data through 2014 found the same pattern, with even stronger estimated effects. The guests argue the main policy lesson is not abortion politics alone, but broader efforts to reduce unwanted pregnancies and support children.

Key Arguments: Legalized abortion may have reduced crime by decreasing the number of children born into circumstances associated with higher later offending. The mechanism is indirect: unwanted children are more likely to experience worse health, education, and behavioral outcomes, which can increase criminal risk. Because abortion legalization varied across states and time, researchers used a collage of imperfect natural experiments rather than randomized trials. The original paper’s correction weakened but did not overturn the conclusion; the updated analysis made the estimated relationship even stronger. The decline in crime likely had multiple causes, including increased policing, imprisonment, and the fall of crack-cocaine violence, plus lead removal and possibly abortion. The episode argues that public debate is too often framed as single-cause storytelling instead of multicausal analysis. Policy relevance is limited for abortion beliefs themselves, but potentially important for crime policy and child-wellbeing policy. Reducing unwanted pregnancies on the front end is presented as a preferable long-term strategy to simply managing downstream consequences.

Data Points: Supreme Court draft status: Authentic but not final - Leaked draft ruling discussed at the start of the episode Roe v. Wade decision year: 1973 - The landmark ruling that legalized abortion nationwide States that had legalized abortion before Roe: 5 states - New York, California, Washington State, Alaska, and Hawaii Democrats favoring legal abortion: 80% - Pew Research Center figure cited in the transcript Republicans favoring legal abortion: just over 35% - Pew Research Center figure cited in the transcript U.S. abortions at peak: 1.5 million per year - Approximate annual abortions at the high point after legalization U.S. live births at peak comparison: roughly 4 million live births - Used to contextualize the abortion volume Abortions per 1,000 live births at peak: 345 - Magnitude used in Levitt and Donahue’s back-of-the-envelope calculations Violent crime decline in the 1990s: more than 30% from 1991 to 2001 - National drop cited by Levitt New York City home-invasion/robbery example: more than 2,200 in 1990 vs fewer than 500 in 2021 - Illustrative local example of the crime decline Estimated share of crime drop explained in original paper: as much as 50% - Conclusion of the 2001 Quarterly Journal of Economics paper Crime difference across abortion-rate states: roughly 30% by 1997 - Difference between highest- and lowest-abortion-rate states in the original analysis Updated relative crime decline: about 60% more - High-abortion-rate states versus low-abortion-rate states from 1997 to 2014 Estimated total share of long-run crime decline: 80% to 90% - Levitt’s characterization of the cumulative effect implied by the updated analysis Current annual U.S. abortions: about 640,000 - Compared with about 1.5 million at the peak

Pivotal Quotes: "“egregiously wrong from the start.”" — Justice Samuel Alito: Quoted description of the leaked draft opinion on Roe v. Wade "“legalized abortion might have reduced crime”" — Steve Levitt / John Donahue: The core hypothesis that sparked the research "“unwantedness is super powerful.”" — Steve Levitt: Levitt reflecting on what the research suggests about children’s outcomes and parenting

Implications: The episode argues that complex social trends often have multiple causes and that abortion restrictions could have downstream crime effects, though abortion policy should not be decided on that basis alone. It also suggests policymakers should focus more on preventing unwanted pregnancies and supporting wanted, well-cared-for children.