Static-99 and Static-99R May NOT Work as Well for Latino Offenders

The Static-99 is the most widely used sex offender risk assessment measure in the world, but predictive effects for the Static-99 have been relatively low in several U.S. samples, particularly in states with high levels of racial or ethnic diversity. The current study uses data from one of the largest predictive validity studies of Static-99 in the U.S. to examine whether Static-99 and Static-99R scores predict recidivism similarly for White (n = 912), Black (n = 411), and Latino (n = 588) offenders. Although predictive effects with both measures were, at times, large enough to reach statistical significance for White and Black offenders, they were never large enough to reach significance for Latino offenders. On the Static-99R, the measure recommended for use in practice, AUC values ranged from .62 to .69 (d = .43 to .71) for Black offenders, .59 to .65 (d = .33 to .56) for White offenders, but only .56 to .58 (d = .17 to .33) for Latino offenders. Findings have implications for fairness in testing and highlight the need for continued research regarding the potentially moderating role of offender race/ethnicity in risk research.

Static-99, Static-99R, sex offenders, racial diversity, ethnic diversity, recidivism

The predictive validity of Static-99 and Static-99R total scores was examined for a sample of 1911 offenders released from prison in Texas between 1999 and 2004. The purpose was to examine whether predictive effects were similar for offenders from different ethnic or racial backgrounds. The total sample was divided into three subsamples for the purposes of these analyses: Black offenders (n = 411), Latino offenders (n = 588), and White offenders (n = 912).

“The Static-99 is a 10-item actuarial risk assessment measure designed to assess sexual recidivism risk. Items include: young age at release, prior convictions for non-sexual violence, number of prior sentencing dates, history of offending against unrelated victims, history of offending against male victims, and number of prior offenses. Each of the items except number of prior offenses is scored as 1 (present) or 0 (absent). Scores for the number prior offenses item range from 0 (none) to 3 (four or more convictions, or six or more arrest charges). Scores on the Static-99 can range from 0 to 12, with higher scores indicating greater risk of sexual re-offense. The Static-99R is identical to the Static-99 with the exception of scoring rules for the age at release item. On the Static-99, offenders who were under the age of 25 receive a score of 1, and those 256 years or older receive a score of zero. On the Static-99R, age is scored so that those aged 18.0 to 34.9 years receive a score of 1, those aged 35.0 to 39.9 receive a score of 0, and those aged 40.0 to 59.9 receive a score of -1, and those aged 60.0 years or older receive a score of -3. As a result, Static-99R scores can range from -3 to +12” (p. 234).

The Static-99 (Mean = 2.94, SD = 1.86) was scored for each offender approximately 16 months prior to release as part of a standard process to identity those offenders who might be civilly committed as sexually violent predators. In addition, Static-99 scores were transformed into Static-99R scores (Mean = 2.37, SD = 2.24) by using offenders’ birth and release dates.

Results indicated that Black offenders received significantly higher Static-99 scores than White offenders, and both Black and White offenders received significantly higher scores than Latino offenders. On the Static-99R, Black offenders received significantly higher scores than both White and Latino offenders.

“Neither the Static-99 nor Static-99R scores were predictive of recidivism among offenders released under mandatory supervision” (p. 235). However, both the Static-99 and Static-99R showed statistically significant predictive validity for violent or sexually violent recidivism among discharged offenders.

“There was no instance in which scores on the Static-99 or Static-99R were statistically significant predictors of sexually violent recidivism…In terms of absolute value, predictive effects were always largest among Black offenders and almost always smallest among Latino offenders. Effects tended to be larger for the Static-99R than Static-99 scores, although these differences were typically small” (p. 235). Predictive effects for the Static-99 and Static-99R were higher for the broader category of violent or sexually violent recidivism. “Once again, in terms of absolute value, effects were largest among Black offenders and smallest among Latino offenders. In the overall sample, Static-99R scores were statistically significant predictors of violent or sexually violent recidivism among Black and White offenders, but not Latino offenders” (p. 235).

“Our overall conclusion from these findings is that the Static-99 and Static-99R may not work as well for Latino offenders as White and Black offenders” (p. 238).

“We observed a pattern of differences in the performance of the Static-99 measures across racial and ethnic groups, with the weakest effects for Latino offenders. But the crucial question is whether these differences are significant and meaningful. On the one hand, our univariate findings seem to suggest that both versions worked poorly for Latino offenders. Indeed, neither the Static-99 nor the Static-99R was a statistically significant predictor of recidivism among Latino offenders, while scores on both measures were, at times, statistically significant predictors of recidivism among White and Black offenders. In other words, these measures were better-than-chance predictors among White and Black offenders, but not among Latino offenders. On the other hand, the difference in predictive effects between Latino and other offenders in this study was not large enough to reach statistical significance. In other words, predictive effects were statistically significant for discharged White and Black offenders, but they were not necessarily significantly larger than those for Latino offenders” (p. 238).

The authors conclude that the appropriateness of the use of the Static-99 instruments with Latino offenders has not been clearly established and indicate that it appears “premature to assume that the Static-99 measures are equally predictive across Black, White, and Latino offenders, at least in the United States” (p. 238).

In terms of what might account for the lower scores on the Static-99 measures among Latino offenders, the authors note that scores may be attenuated because of missing data, measurement error in recidivism might occur as a result of differences in rates of reporting of certain types of crimes committed by Latino offenders (e.g., cultural values and expectations that influence sexual offense reporting), and functional inequivalence of the Static-99 items across cultures (e.g., differences in culturally-relevant values among sec offenders of varying ethnicities and race” (p. 239).

“Practitioners who evaluate sexual offenders might feel uncomfortable basing conclusions about a Latino offender on the results of a measure that is not significantly related to recidivism among Latino offenders…These findings highlight the importance of acknowledging that Latino offenders were not included in the research studies used to develop Static-99 items, and that there appear to be relatively few Latino offenders included in the 2003 or 2009 norms” (p. 239).

“Future research should also examine the utility of specific Static-99 and Static-99R items across racial and ethnic groups, especially among offenders who immigrated to the United States. The Static-99 and Static-99R are scored using historical information; the accuracy of scores likely depends on the accuracy of the historical information available to those assigning scores. Research that takes into account the availability (or lack thereof) of collateral data for foreign-born offenders may help shed light on this issue. Item-level research can also examine the functional equivalence or inequivalence of sexual recidivism predictors (e.g., items) across cultures. While research suggests these items are relevant to re-offending among White offenders in Canada, Europe, and the United States, they may be less relevant (or carry different implications) among Latino offenders, especially immigrant offenders. In addition, length of time in the U.S. and acculturation may moderate the utility of the Static-99 items (as well as other actuarial measures). That is, these items may be better predictors among those with higher levels of acculturation to the United States” (p. 240).

As always, please join the discussion below if you have thoughts or comments to add!