Effects of Adult Prison vs. Youth Custody/Detention Centers by Type of Offense (in the Next Eight Years)
| Independent Variable: Adult Prison | ||||||
|---|---|---|---|---|---|---|
| 1963 | 1975 | |||||
| Softer Treatment | Harsher Treatment | |||||
| 365 Days (1) | 274 Days (2) | 183 Days (3) | 365 Days (4) | 274 Days (5) | 183 Days (6) | |
| Thefts | −0.999** (0.504) | −1.163** (0.513) | −1.106* (0.582) | 0.951* (0.512) | 1.175** (0.578) | 1.350** (0.681) |
| Violent offenses | −0.817** (0.327) | −0.802** (0.379) | −0.972* (0.503) | 0.100 (0.072) | 0.143* (0.083) | 0.209** (0.103) |
| Sexual offenses | −0.028 (0.034) | −0.021 (0.029) | −0.011 (0.010) | 0.003 (0.019) | 0.001 (0.020) | −0.011 (0.026) |
| Burglaries/robberies | −0.349 (0.267) | −0.346 (0.286) | −0.173 (0.377) | −0.053 (0.040) | −0.065 (0.047) | −0.104* (0.057) |
| Minor offenses | −0.381 (0.333) | −0.315 (0.390) | −0.351 (0.560) | |||
| Frauds | −0.050 (0.229) | 0.115 (0.236) | 0.332 (0.251) | 0.042 (0.092) | 0.077 (0.104) | 0.051 (0.123) |
| Criminal damage | −0.216*** 0.083) | −0.155* (0.091) | −0.101 (0.120) | 0.084* (0.050) | 0.087 (0.058) | 0.098 (0.069) |
| Drug offenses | 0.166 (0.118) | 0.155 (0.129) | 0.172 (0.156) | 0.101 (0.129) | 0.172 (0.147) | 0.293* (0.177) |
| Motoring offenses | −0.051 (0.100) | −0.095 (0.090) | −0.160** (0.065) | |||
| Other offensesa | −0.334** (0.148) | −0.337** (0.155) | −0.371** (0.189) | 0.651*** (0.235) | 0.784*** (0.266) | 0.846*** (0.317) |
| Observations | 558 | 457 | 288 | 5,440 | 4,110 | 2,767 |
Notes: Columns 1–3 report the effects of experiencing prison rather than youth custody/detention centers for the 1963 cohort of the Offenders Index Cohort Data (Home Office Research, Development and Statistics Directorate) over the next eight years. Columns 4–6 report the effects of experiencing prison rather than young offender institutions for the 1975 cohort of the Police National Computer (Ministry of Justice, PNC/Criminal Histories Team Justice Statistics Analytical Services) over the next eight years. The estimation is conducted through a local linear regression constructed with a triangular kernel regression. Each column corresponds to a different bandwidth selection: in Columns 1 and 4 the bandwidth is 365 days, in Columns 2 and 5 the bandwidth is 274 days, and in Columns 3 and 6 it is 183 days. Robust standard errors are reported in parentheses:
↵* p < 0.1,
↵** p < 0.05,
↵*** p < 0.01.
↵aOther offenses include mainly: failing to surrender to bail, going equipped for stealing, and other offenses against the state or public order.