Age-Crime Curve Modeling
Also known as: Age-Crime Relationship Modeling, Age-Offending Curve, Aggregate Age-Crime Distribution, Crime-Age Profile Modeling
Age-crime curve modeling fits statistical functions to the well-known relationship between age and offending: crime rises sharply in adolescence, peaks in the late teens or early twenties, and declines through adulthood. Brought to prominence by Hirschi and Gottfredson's 1983 claim that this curve is invariant, and elaborated by Farrington, the modeling task is to capture its characteristic skewed, single-peaked shape and to debate what it implies about the causes of crime.
Key highlights
- Captures one of criminology's most robust empirical regularities in a compact, comparable functional form.
- Peak age and skew summaries allow comparison of the age-crime relationship across crime types, eras, and populations.
- Count-regression formulations correctly handle the skewed, over-dispersed nature of offending counts.
- Provides the descriptive backdrop against which developmental theories and trajectory models are tested.
- Useful for age-adjustment and forecasting crime given a population's age structure.
Intuition
This section is available to Pro members. Upgrade to Pro
How it works
This section is available to Pro members. Upgrade to Pro
When to use it
Use age-crime curve modeling when you have age-specific offending data and want to describe, summarize, or compare the age distribution of crime — its peak age, shape, and decline — across offense types, periods, places, or cohorts. It is foundational for developmental and life-course criminology and for adjusting analyses for age composition. It is descriptive at the aggregate level: it characterizes the population curve but cannot, by itself, distinguish whether the shape arises from changing participation or changing individual frequency, a question that requires longitudinal, individual-level data and trajectory or count models. It is also sensitive to data quality, age-range truncation, and cohort/period confounding.
Strengths & limitations
- Captures one of criminology's most robust empirical regularities in a compact, comparable functional form.
- Peak age and skew summaries allow comparison of the age-crime relationship across crime types, eras, and populations.
- Count-regression formulations correctly handle the skewed, over-dispersed nature of offending counts.
- Provides the descriptive backdrop against which developmental theories and trajectory models are tested.
- Useful for age-adjustment and forecasting crime given a population's age structure.
- Aggregate curves cannot reveal whether the shape reflects participation, frequency, or both — the central interpretive limit.
- The 'invariance' claim is contested: peak age and decline rate do vary by offense type, sex, cohort, and context.
- Cross-sectional curves confound age, period, and cohort effects, which cannot all be separated without strong assumptions.
- Polynomial fits can oscillate or behave implausibly at the youngest and oldest ages, distorting peak and tail estimates.
- Official-record curves reflect detection and justice processing as well as actual offending.
Common pitfalls
This section is available to Pro members. Upgrade to Pro
Applications
This section is available to Pro members. Upgrade to Pro
Frequently asked
Why does the age-crime curve peak in late adolescence?
Explanations differ by theoretical camp. Developmental accounts point to peaking peer influence, identity formation, and limited stakes in conformity during adolescence, followed by adult bonds (work, marriage) that promote desistance. Hirschi and Gottfredson argued the decline reflects an age effect that no social variable fully explains. The aggregate curve alone cannot adjudicate these, which is why individual-level data are needed.
Is the age-crime curve really invariant?
It is remarkably robust in gross shape — a single adolescent peak with a right-skewed decline appears almost everywhere — but it is not strictly invariant. The peak age, height, and rate of decline differ by offense type (property crime tends to peak earlier than violence), by sex, and across cohorts and eras. The invariance claim is best read as a strong regularity with systematic, theoretically meaningful exceptions.
Why can't the aggregate curve settle the criminal-career debate?
The same aggregate age-crime curve can arise from different individual-level processes: more people offending during adolescence (a participation effect) or active offenders offending more often when young (a frequency effect), or any mix. Distinguishing these requires longitudinal, individual offending histories analyzed with trajectory and count models — which is exactly why the criminal career paradigm pushed the field toward individual-level data.
Sources
- 1.Hirschi, T., & Gottfredson, M. (1983). Age and the explanation of crime. American Journal of Sociology, 89(3), 552–584.
- 2.Farrington, D. P. (1986). Age and crime. Crime and Justice, 7, 189–250.
You have read it. What now?
Cite this page
ScholarGate. (2026, June 22). Age-Crime Curve Modeling. ScholarGate. https://scholargate.app/criminology/age-crime-curve-modeling