In 2010, Nagy's pretrial services staff screened more than 40,000 defendants in the Travis County jail, to determine who might be eligible for a personal bond. Only a discrete class of defendants is automatically ineligible – including those charged with capital crimes and those who've violated parole or forfeited a previous bond. In all, roughly 32,000 of those screened made it to a full-fledged pretrial investigation, for which staff collects references, conducts a background check, and considers a variety of issues – including employment status and family support – that might suggest whether a defendant will show for court or be rearrested. The agency then makes a recommendation to the courts about whether the person should be granted a personal bond. In 2010, 63% of defendants released pretrial were released on personal bond. That same year the agency posted a court failure-to-appear rate of 13% and a rearrest rate of 14%, according to annual reports the agency is required to file. In contrast, that same year defendants released on commercial bond failed to appear in court 20% of the time, according to information collected by Travis County.Attorneys for bail bond companies, who were quoted extensively in the story, claimed their failure to appear (FTA) rate was lower - more like 2% compared to the county's 13%. By contrast, Smith reported, Travis County calculated that clients of commercial bail companies posted failure to appear rates of 20% and 17% in 2010 and 2011, respectively.
Nagy believes the FTA and rearrest numbers will shrink with the introduction in November of a research-based risk-assessment tool, designed and validated specifically for Travis County defendants, in order to help her staff more objectively determine a person's likelihood of success if released on personal bond. "That's really consistent with all the research on decision-making. ... These tools do significantly better than just the person" making a decision, Nagy says. "Making a pretrial decision without risk assessment is like playing golf in the fog."
From a mathematical perspective, it makes sense to me that, used widely along with a validated risk assessment, personal bonds would have lower FTA rates. In general, most people will just show up for court, especially if pretrial services supports that tendency with a multimedia program of reminders (phone calls, texts, emails, snail mail, etc.). If the risk assessment can accurately identify folks who are likely to show up anyway, then the bail-bond industry's client pool would consist of people judged by pretrial services to pose a higher risk of flight. Thus, if used widely, you'd expect the FTA rate for bail bond clients - almost by definition, if the risk-assessment instrument is doing its job - to be higher than the lower-risk folks judged eligible for release on their own recognizance. But until Travis County began issuing personal bonds for the majority of defendants (63% in 2010) released pretrial, creating a large enough pool to make a valid comparison, one could only speculate: Someone had to test the theory and document the results. Now Nagy has demonstrated it can work.
One also can see outlines of potential legislative reforms suggested in the story, at a minimum requiring bail-bond agents to produce reporting and documentation on their clients' outcomes comparable to the records published by pretrial services. I've been around these blocks for several years and agree with Nagy that the lack of reporting on the commercial bail side often frustrates any serious mathematical analysis of county-level outcomes. This is a long-standing issue affecting many counties, and generating public data sufficient to make apples to apples comparisons between bail-bond companies and pretrial services could be an important first step toward shifting the debate from the sort of animated hype and misinformation put out by the bail-bond attorneys quoted in the story to a discussion animated by data and outcomes. Hype aside, what works? Is there data to prove it? Nagy's suggestion to require bail agents to submit reporting comparable to reporting requirements for pretrial services is to me a stroke of genius: That would resolve a lot of the confusion surrounding the elusive apples to apples outcomes for pretrial services and commercial bail. ("'Data data, data,' he cried impatiently, 'I cannot make bricks without clay.'")
Excellent reporting from Jordan Smith (read the whole thing), and I hope Dr. Nagy succeeds in making Travis' expanded use of personal bonds a model for its sister jurisdictions around the state. Folks in other Texas counties should pay heed: When administered through a validated risk assessment, personal bonds don't reduce the likelihood defendants appear in court - which after all is the purpose of bail - and save huge sums in reduced incarceration costs at county jails.
RELATED: See the Justice Policy Institute's latest series of reports on the U.S. bail system.