BOP Update: Mass Overincarceration And Miscalculation Of Good Time

This morning Professor Berman has us thinking about the huge material and human costs of overincarceration in America (here and here). Which leads me to thoughts once again of the Bureau of Prisons policy of shorting federal prisoners of seven days a year of good time credit. When constructing the sentencing guidelines, the Sentencing Commission created the Sentencing Table -- the graph upon which every federal prison sentence is charted -- assuming that prisoners were required to serve 85% of the sentence imposed and could therefore receive up to 15% in good time. Nevertheless, the BOP has only allowed a maximum of 12.8% of good time credits against the sentence imposed, thereby adding hundreds of millions of prison cost dollars and over 34,000 years of overcarceration to the sentences intended by Congress and the Sentencing Commission.

The story starts in 1987, when the Sentencing Commission’s staff was assigned the task of creating a baseline for the Sentencing Table, upon which all federal sentences were to be graphed. To create the Sentencing Table, Sentencing Commission staff collected a large sample of sentences for a broad array of crimes and determined the actual time served as a baseline. United States Sentencing Commission, Supplemental Report On The Initial Sentencing Guidelines And Policy Statements (June 18, 1987) at 23. Then, the Commission "adjusted for good time" by figuring out the longer sentence for which the actual time served would be 85%:

"Prison time was increased by dividing by 0.85 good time when the term exceeded twelve months. This adjustment corrected for the good time (resulting in early release) that would be earned under the Guidelines. This adjustment made sentences in the Levels Table comparable with those in the Guidelines (which refer to sentences prior to the awarding of good time)."

Id.; see also U.S.S.G. Ch.1, Pt. A, § 3, para. 3 (2005) at 9 ("Honesty is easy to achieve: The abolition of parole makes the sentence imposed by the court the sentence the offender will serve, less approximately fifteen percent for good behavior.").

Thus, every federal prisoner has had a term of imprisonment imposed based on a Sentencing Table that assumes good time credit based on 15% of the sentence imposed. But the BOP takes a different view. The BOP does not base good time on the term of imprisonment, but substitutes a "time served" formula that reduces maximum good time credit by seven days for every year of the sentence imposed. The BOP formula requires that ideal prisoners serve at least 87.2% of the sentence imposed. For example, on a year-and-a-day sentence, maximum good time credit is 47 days, not 54 days; on a 60-month sentence, the maximum good time credit is 235 days, instead of 270 days; on a 120-month sentence, the maximum good time credit 470 days, not 540 days. Until the BOP changes its method of calculation to mirror the method upon which the Sentencing Table is calibrated, every bottom-of-the-guideline sentence is 2.2% higher than the Sentencing Commission intended based on its statistical methodology.

The over-incarceration multiplies with every added year of the sentence. For all federal prisoners eligible for good time, the total time involved is over 34,000 years (188,410 prisoners x 7 days a year x 9.5 average sentence over a year and less than life ÷ 365 days in a year = 34,326 years). At $22,265.00 per year for non-capital incarceration expenditures, this amounts to over $764 million in taxpayer money that Congress did not intend or authorize to expend on incarceration for current prisoners, and over $66 million more for each new year.

The circuits are in disarray on reasoning but relatively consistent in result: the good time statute is ambiguous, so instead of applying the rule of lenity, the courts have deferred to the BOP’s severe construction. Three district courts have ruled our way in well-reasoned opinions that have been overturned by circuit courts. Moreland v. Fed. Bureau of Prisons, 363 F. Supp. 2d 883 (S. D. Tex. 2005); Williams v. DeWalt, 351 F. Supp. 2d 412 (D. Md. 2004); White v. Scibana, 314 F. Supp. 2d 834 (W.D. Wisc. 2004). This past spring, the Supreme Court glanced at the issue in Moreland, drawing a shocking response from Justice Stevens: the Moreland district court appears to be correct, but, in the absence of a circuit split, certiorari would be denied!

How do Defenders respond to Justice Stevens? In his statement accompanying a denial of certiorari on the good time issue (available here), Justice Stevens said we appear to be right that the statute calls for 54 days credit for every year of the sentence imposed: "[B]oth the text and the history of the statute strongly suggest that it was not intended to alter the pre-existing approach of calculating good-time credit based on the sentence imposed." Then, instead of granting certiorari, he encourages further litigation in the absence of a circuit split: "[T]he question has sufficient importance to merit further study, not only by judges but by other Government officials as well."

Two courses of action seem reasonable, both of which involve continued and renewed litigation in the district courts. First, and most simply, we must continue giving the Supreme Court the opportunity to resolve the BOP's misconstruction of the good time statute (as in this sample cert petition). This is a question of exceptional importance: no other criminal justice issue has a fraction of the potential effect on the federal prison population. The Court should recognize that the reason the Circuit Court opinions are as scattered and as illogical as they are may be because of the administrative consequences of a split in the Circuits: uniformity is desirable in the administration of sentences. Only the Supreme Court, by correctly interpreting the statute, can compel the BOP to resolve the problem with a simple computerized adjustment for uniform recalibration of good time credits.

The second approach is relitigating under section 706 of the Administrative Procedure Act, a theory that presents a question of first impression. Under section 706 and the Supreme Court decision in State Farm, agency action is invalid if it is arbitrary, capricious, unreasonable, or an abuse of discretion under 5 U.S.C. § 706. The institutionalized skewing of sentences to add actual time to Guidelines sentences meets all the § 706 factors for unlawful agency action, especially given the statutory call for a sentence sufficient but not greater than necessary to serve the purposes of sentencing. 18 U.S.C. § 3553(a). It makes no sense to require service of time 2.2% greater than the Sentencing Commission itself intended based on the BOP’s refusal to correspond its interpretation of an ambiguous statute to the Sentencing Commission’s interpretation (never mind the violation of the rule of lenity).

And the APA strategy has worked in the past. When the Supreme Court ruled against us on the BOP’s DAP rules, the Court left open the possibility of an APA challenge in footnote 6. Lopez v. Davis, 531 U.S. 230 (2001). The Ninth Circuit later invalidated those same BOP rules for violation of the notice-and-comment provisions of the APA in Paulsen v. Daniels, 413 F.3d 999 (9th Cir. 2005).

Similarly, we should be raising this different APA challenge, for which there is no adverse precedent, in every jurisdiction where prisoners are serving sentences. We should preserve the statutory issues, especially the anomaly of Chevron deference to the Executive Branch on an ambiguous penal statute, and try to get circuits to reconsider en banc. But the next generation of litigation should aim directly at the absurdity of a Sentencing Table constructed on the basis of a statutory construction that the BOP later did not follow, not because there was a considered decision, but because the BOP misinterpreted the statute to unambiguously require their incorrect construction of the statute.

Steve Sady, Chief Deputy Federal Public Defender, Portland, Oregon