There has been a long-running comparison between California and Texas as to which of the country’s two most populous states — each with a wholly distinct flavor of governing style — best serves its citizens. Commentators may point to this benefit or that, but the direction of net migration by ordinary citizens shows people seem to prefer the Lone Star State.

After a recent hearing in Texas’ House Criminal Jurisprudence committee, yet another point of contrast can be added to the discussion: the issue of pretrial bail reform.

Although pretrial jail populations don’t always receive the attention other areas of policy do, their recent expansive growth has become difficult to ignore. In one estimate, America’s pretrial jail population grew from about 83,000 detainees in 1970 to roughly 462,000 detainees by 2013, a five-fold increase. Texas has experienced similar growth. In 1994, almost 33 percent of the jail population was comprised of pretrial detainees, better than the national average that year (about 50 percent). By the beginning of this August, however, that had jumped to almost 76 percent.

It bears remembering that these people, while suspected of crimes, have not been convicted, meaning they are still presumed innocent. Continuing to detain these people, many of whom are low-risk, is an expensive proposition, not just in raw dollars, but in terms of liberty. It may also run the risk of actually increasing the likelihood of future lawbreaking, according to a 2013 study from the Arnold Foundation.

Enter bail reform. Several states, most recently California, have begun moving away from cash bail in favor of using risk-based systems. Critics of the status quo decry cash bail’s propensity to disproportionately affect defendants who may be detained solely on the basis of wealth, instead of their risk to the public. By making release decisions based on an objective evaluation of risk, using validated risk-assessment tools, more defendants ought to be released safely pending trial.

While California’s efforts signal the necessity of developing more equitable and risk-sensitive pretrial practices, they nonetheless miss the mark. Worse, some elements threaten to overcorrect and further embed problems that lawmakers sought to alleviate.

For starters, California has effectively eliminated their cash bail system and adopted a more or less release-or-no-release practice. Basing release decisions on risk is a good idea and supported by data, but the original purpose of bail was to provide a surety of good conduct upon release, including appearance in court. Eliminating all financial bail reduces a court’s power to ensure this and removes a viable arrow from a judge’s quiver. Risk-assessment tools may do a better job of predicting future criminal offense, but cash sureties remain an effective means of getting people to show up for their court date.

Judges should be able to use the benefits of both tools. California has shut the door on that possibility.

While California has stripped discretion in this sense, they’ve expanded it elsewhere, perhaps too much. California’s new law grants broad power to preventatively detain defendants suspected of a broad range of offenses. While this is unobjectionable for violent or repeat offenders, unjustified risk aversion may end up capturing lower-risk people and diminish any intended progress.

Fortunately, Texas has taken a more sensible approach to reform. A wide array of stakeholders, including Chief Justice Nathan Hecht of Texas’ Supreme Court, discussed with committee lawmakers a proposal for using risk assessments to better inform a judge’s decisions, but importantly, retain all current forms of bail. Research by Texas A&M professors shows that this would more effectively prevent misclassifications: either detaining low-risk people who can be released safely, or releasing dangerous people who ought to be detained before trial.

In Kentucky, the rate of new offenses by those released fell by almost 15 percent when they based release on risk. Additionally, the ability to preventatively detain risky defendants ought to be expanded, but not so much as to potentially foreclose any defendant from being released pending trial.

We should be expanding options, not taking from one hand to give to the other.

No pretrial system will ever be perfect. However, Texas is again showing that it approaches policy reform from a more even keel than our coastal rival. Pretrial release should be based primarily on risk, not the shake of a Magic-8 ball, but judges should retain tools that they already have.