Table 5.

Effects of Any Maltreatment on Criminal Activity—Fixed Effects Estimates with Twin and Sibling Samples

(Using > 10X cutoff for physical abuse and being left alone when should have been supervised. Each coefficient is from a separate regression.
	Twin Fixed Effects	Sibling Fixed Effects	Sibling Fixed Effects Males	Sibling Fixed Effects Females
Any nondrug	0.144*** (0.045)	0.086*** (0.030)	0.119* (0.063)	0.110** (0.053)
Armed robbery	0.011 (0.017)	0.008 (0.013)	0.034 (0.028)	0.026 (0.016)
Burglary	0.047*** (0.014)	0.018* (0.010)	0.011 (0.022)	0.052** (0.022)
Damaged property	0.089** (0.036)	0.058** (0.024)	0.081 (0.053)	0.057 (0.035)
Assault	0.067** (0.031)	0.042** (0.021)	0.064 (0.046)	0.002 (0.031)
Theft > $50	0.039* (0.020)	0.006 (0.016)	−0.011 (0.030)	0.026 (0.024)
Any hard Drug	0.083** (0.034)	0.005 (0.024)	0.040 (0.048)	−0.021 (0.035)
Crime victim	0.080** (0.037)	0.066** (0.025)	0.061 (0.051)	0.051 (0.038)
Ever convicted	0.023 (0.031)	−0.006 (0.021)	−0.038 (0.046)	0.009 (0.022)

Notes: Robust standard errors in parentheses.

A *, **, *** indicates significance at 90 percent, 95 percent, 99 percent respectively.

The sample sizes range from 926 to 928 in Column 1, from 2,211 to 2,115 in Column 2, from 655 to 666 in Column 3, and from 828 to 830 in Column 4. Twin fixed effects models include child gender and indicators for very low birth weight, low birth weight, ADHD, and parental reports about whether the child was bad tempered and had learning problems as of Wave I. Sibling fixed effects models also include indicators for age.