Last Fall, the House considered a “three-strikes” or “habitual offender” bill. Both branches have now passed versions of the bill and the bill is presently before a conference committee.
I voted against the bill based on concerns including the likely prison impact — see One Tough Vote. Data continue to emerge that reinforce concerns that the prison population could rise substantially.
Massachusetts already has a law that requires imposition of the maximum available sentence for a new offense committed by a person twice previously convicted of a serious offense. A serious offense is defined as one resulting in at least a three-year committed state prison sentence. Both the House and Senate versions of the bill expand the base of offenders subject to maximum sentences and increase the maximum sentences available.
The Massachusetts Sentencing Commission recently released a study of the respective versions of the bill as passed by the House and Senate.
In two respects, the MSC analysis would appear to diminish concerns about the impact of the bill — it shows that the number of offenders annually eligible for habitual offender treatment would not increase too radically, only 20% in the House version. It also confirms that the number of inmates actually sentenced under the habitual offender law is low (fewer than 10 per year).
However, the MSC analysis shows that the annual number of offenders eligible for habitual offender sentencing — whether or not actually sentenced as habitual offenders — is much greater than previous official estimates would suggest. The MSC data suggest that many more habitual offender cases are settled by plea bargains in which the habitual offender count is dropped. While only 4 offenders were sentenced as habitual offenders in 2009, 235 offenders were eligible. The other 231 do not appear as habitual offender cases in the Department of Public Safety data, which only reflect the counts upon which the current sentence is actually imposed.
Most criminal cases do settle through plea bargains. In the Superior Court in 2011, there were 3,891 pleas but only 683 trials; the trial count includes acquittals, so over 85% of convictions were by plea.
In most criminal cases, the defendant is charged with several different counts — perhaps several specific offenses and an additional habitual offender count. To accommodate any workable plea bargain, the habitual offender count must be dropped; no discount is possible if the habitual offender count is not dropped, because it requires imposition of the maximum sentence. Note that the habitual offender count can be dropped even after trial: Habitual offenders are entitled to a bifurcated trial process in which the Commonwealth must first convict on the third strike and then in a second trial prove that the offender is in fact the offender who committed the first two strikes. A sentencing agreement can entered even after the first trial.
The Sentencing Commission, although it tracks conviction history, does not track charges that do not result in a conviction, so it is not possible (based on their data) to make a completely authoritative statement about how indictment decisions and the plea bargaining process interact to reduce the 235 eligible offenders down to only 4 so-sentenced habitual offenders. But given the overwhelming prevalence of plea bargaining in the courts and the special features of habitual offender cases, it is reasonable to believe that the most of eligible offenders are so charged and that the habitual count is dropped in an agreement either before or after trial on the current offense.
A new habitual offender law would change sentencing outcomes even in the cases where habitual offender charges are dropped. In any bargaining process, the outcome is shaped by the “best alternative to a negotiated agreement”. So, lengthening of the mandatory sentences for habitual offenders will increase the length of sentences agreed to when habitual offender counts are dropped. According to the Sentencing Commission among the 235 people eligible for existing-law habitual offender treatment in 2009, the average sentence was 63 months, while it would have been 99 months if they had all been so-convicted and subjected to the existing mandatory. This suggests a going-rate plea-bargaining discount of roughly 37% for these cases.
Future predictions about sentencing behavior are necessarily uncertain, but available data and experience support the following estimating assumptions: (a) that most of the eligible habitual offenders will be so-charged and (b) that the habitual offender count will usually be dropped, and (c) that actual sentences will be materially below sentencing exposure as under the applicable habitual offender law — assume the known going discount of 37%.
Stock and flow analysis translates estimates of annual prison commitments into estimates of steady-state prison population. For example, life without parole equates to roughly a 40 year sentence for a typical offender in his 30’s. If 10 people per year are sentenced to a 40 year sentence, the resulting prison population 40 years later (when they are mostly still there) is 400.
In the table below, the steady-state prison impact of the two versions of the habitual offender bill is estimated (from the Sentencing Commission’s data) by multiplying the average increase in sentence length for offenders by the number of offenders affected. The numbers shown reflect exposure — the sentence length if the habitual offender were so-sentenced — and assume that the change in exposure roughly predicts the change in agreed sentence when the habitual offender charge is dropped in plea bargaining.
Change in Sentencing Exposure for Habitual Offenders
Minimum Parole Eligibility (MPE) under current law and revisions if all habitual offenders sentenced as habitual offenders
|Count||Current Law -- Average Months to MPE||Revised Law -- Average Months to MPE||Current Law -- Annual Man Years||Revised Law -- Annual Man Years||Increase in Man Years of Exposure||Discounted Increase -- Likely Population Increase|
|Affected under H. 3818||281||91||156||2138||3658||1520||963|
|Affected under S.2080||307||96||181||2462||4631||2168||1,374|
Note that a key additional assumption in the estimate is that prisoners are released at their minimum parole eligibility (MPE), or at least that they serve a sentence roughly proportional to their MPE. Of course, the future behavior of the parole board is hard to estimate with confidence.
The bottom line increase would certainly be below the undiscounted estimate of 2,168 under the Senate version, but could easily approach the discounted increase of 1,374. It is not unreasonable to fear an prison population increase of over 1,000 inmates in the Senate version of the bill — over a 10% increase. (Click here for the Department of Corrections Annual Report and see page 38 for a recent inmate count.)