Most Frequently Asked Questions Prison & Sentencing Impact Assessments

This document provides answers to the most frequently asked questions about the Commission's prison and sentencing impact assessments. (Printable PDF version.)

Answers

  1. What is a prison and sentencing impact assessment?

    A prison and sentencing impact assessment is the Commission's appraisal of the impact that changes to the sentencing guidelines or criminal statutes will have on the federal prison population. Commission staff collects data from documents sent by the courts for each case in which a defendant is sentenced. From that data, a database is created that includes information on guideline application, statutory penalties, and sentences. Commission staff analyzes information from this database to estimate the possible prison and sentencing impact of proposed legislation that may affect criminal statutes and the federal sentencing guidelines.

  2. Under what statutory authority does the Commission perform prison and sentencing impact assessments?

    The Commission is required to consider the impact of changes to the federal sentencing guidelines on the federal prison population under 28 U.S.C. § 994(g) and 18 U.S.C. § 4047.

  3. How does the Commission estimate prison and sentencing impact assessments?

    General Methodology

    The Commission uses a prison and sentencing impact assessment model to create its assessments. In simplified terms, the model recalculates the relevant sentencing computation based on specific changes and compares the recalculated sentence to the existing sentence. For example, to assess a proposed change to the drug amounts that correspond to base offense levels (BOLs) in the Guidelines Manual, the model would first recalculate BOLs according to the proposed changes. Then, the model would reassign any pre-existing adjustments or departures/variances that existed (e.g., application of Chapter Three adjustments). Finally the model "respots" the new sentence in the corrected guideline range to a location equivalent to the current sentence.

    The sentencing impact is based on the difference between the original sentence and the new estimated sentence the model predicts. The prison impact is based on the number of beds that will either need to be added (for sentence increases) or will be freed up early (for sentence decreases).

    Detailed Methodology

    The Commission uses a "micro-simulation" model to perform prison and sentencing impact assessments. Micro-simulation is a technique that looks at individual cases and makes specific changes to each case. The micro-simulation prison and sentencing impact model uses all of the actual guideline application (BOLs, specific offense characteristics (SOCs), Chapter Three adjustments, and criminal history points), plus, if available, information from the proposed guideline or statutory changes. The model then assigns, as applicable, a new offense level, criminal history category, guideline range, and/or statutory range for each case that meets the selection criteria.

    Next, the relationship between the actual sentence imposed and the actual guideline range is determined. This relationship is used to calculate the new sentence maintaining the same, proportional relationship to the new guideline range (both for within range and outside the range cases).

    The model then takes statutory minimums and maximums into account. The statutory minimum and maximum applicable in each case are the legal limitations on the sentence ultimately imposed on an offender. The model accounts for and then applies these applicable statutory limitations on the estimation of the new sentence.

    The model assigns a new sentence to the case based on all of the proposed changes. Only the factor being analyzed (statutory or guideline change) is modified; all remaining applicable guideline factors in the case (BOL, SOCs, Chapter Three adjustments, criminal history points, etc.) that did not change are applied as they were at the original sentencing. The change from the actual to the new estimated sentence is thus limited to only the specific guideline or statutory change being modeled.

    The model assumes that if an offender's guideline range changes, the judge will proportionally re-sentence the person to the same place either within the range or outside of the range (i.e., "re-spot" the sentence in a similar place relative to the new estimated guideline range). For example, if the guideline range for an offender is 30-37 months and the offender was sentenced to 30 months, then he was sentenced at the bottom of the guideline range. If his new guideline range is 37-46 months, then the model would proportionately re-sentence him to 37 months, the bottom of the new guideline range. Likewise, if the guideline range for an offender is 30-37 months and the offender was sentenced to 24 months, then he was sentenced 20% below the guideline range. If the proposed guideline amendment increased his final offense level by two levels, then his new guideline range would be 37-46 months. The model would proportionately re-sentence him to 29.6 months, which is 20.8% below the bottom of the new guideline range).

    Occasionally, some information is not available in sentencing documents that the Commission receives from the courts. Although all five primary documents (Indictment, Plea Agreement (if applicable), Pre-Sentence Report (PSR), Judgment and Commitment Order, and Statement of Reasons) may be received by the Commission, exact information (e.g., an offender's exact drug or loss amount) may not be reported. For example, sometimes all that is reported in the PSR is a range of drug amount instead of an exact amount. This range typically corresponds to the base offense level for drug amounts. In order to estimate new sentences for cases involving this type of missing information, the model randomly assigns a drug weight from within the actual BOL range. Using this methodology to assign specific values based on the available information permits inclusion of cases that otherwise would be excluded from the analysis. This maximizes the number of cases included in the analysis.

  4. What assumptions does the model include and how do the assumptions affect the validity of the estimate?

    The prison and sentencing impact assessment model estimates both a new sentence and an estimate of the new time served. The model operates under the assumption that an offender will receive all of the good time credits available. This assumption is also used in calculating beds to be added (increasing sentences) or saved (decreasing sentences).

    The model also makes the following assumptions specific to calculating estimated sentencing and prison impacts:

    • The sentencing impact portion of the model operates under several assumptions. One assumption is that judges will proportionally re-sentence an offender if the guideline range changes. Sentencing impacts may differ depending upon the accuracy of this assumption, especially under varying degrees of change. Also, new sentences are capped at the offender's predicted life expectancy; the accuracy of the life expectancy prediction will vary by individual. Another assumption, as discussed above, is that in cases involving missing information regarding the exact drug or loss amount in which amounts are randomly assigned, the amounts will be uniformly distributed. Further, the model is less accurate when estimating new sentences based on decreases to guideline ranges in the lower sentencing zones (Zones A to C). Finally, the model does not include any time of alternative confinement, because to include such information, the Commission would be required to make difficult assumptions about what percent of offenders would receive split sentences (as compared to offenders receiving probation or straight prison).

    • The prison impact portion of the model assumes a hypothetical steady state prison system. The model assumes that this hypothetical prison system is in balance, i.e., the number of incoming inmates is exactly equal to the number of discharged inmates. The model operates to measure the effect of statutory or guideline changes by measuring the changes in this "balance." For example, if under the model the new sentences are estimated to increase, then offenders will be released later than they are actually; as a consequence, additional prison beds will be required to restore the balance. If the new sentences are shortened, beds are freed up earlier than expected. Because only a single year of data is used, the offender population for the year establishes the parameters of the "balanced" system. The model does not account for changes in law enforcement action or prosecutorial decision making that may occur over time, nor does the model plan for other statutory or guideline changes.


  5. Can the Commission always perform a prison and sentencing impact assessment on a piece of legislation?

    The Commission is not always able to perform a prison and sentencing impact assessment for a piece of legislation due to the insufficient availability of data. The Commission collects data on all of the cases sentenced under the federal sentencing guidelines from the documents that each court sends to the Commission. There are several instances in which the Commission's data does not allow for performing a prison and sentencing impact assessment:

    • New offenses or new aspects of actual offenses which are not currently punished under federal law. In these circumstances, the Commission has no existing data on these new offenses from which to perform a prison and sentencing impact assessment.

    • Increases to the statutory maximum of an offense. All sentences imposed are at or below the statutory maximum. It is unknown what guideline changes would be contemplated by the Commission in response to an increase in the statutory maximum of an offense. As such, there is no information available that would allow the Commission to perform a sentencing or prison impact assessment if legislation proposed increasing the statutory maximum of an offense or group of offenses.

    • Statutes under which there have been zero or very few cases sentenced. The Commission is not able to model the possible impact of legislation for which there may be no, or very few, cases because there is not enough data to perform an accurate assessment. However, such proposed changes likely would not require a prison impact because the changes would logically have a minimal impact on the system.


  6. Can the Commission always perform a prison and sentencing impact assessment for a federal sentencing guideline amendment?

    The Commission is not always able to perform a prison and sentencing impact assessment for a federal sentencing guideline amendment because of one or more of the following reasons:

    • The Commission collects data on all of the cases sentenced under the federal sentencing guidelines from the documents that each court sends to the Commission. However, sometimes potential changes to the federal sentencing guidelines address new offenses or new aspects of actual offenses which are not currently punished under the guidelines. In these circumstances, the Commission does not have the data necessary to perform a prison and sentencing impact assessment. For example, if a proposed guideline amendment increases penalties if the offender is engaged in certain conduct not now accounted for by the guidelines, the Commission's existing data would not include such information because it is not currently relevant under the guidelines in determining the offender's sentence.

    • Occasionally, only a very small number of offenders will be affected by a proposed guideline amendment. Under these circumstances (typically if there are fewer than 50 cases that will be affected), the Commission does not perform a prison and sentencing impact assessment because there is insufficient data available to perform an accurate assessment and any potential impact of the proposed chance would be minimal.


  7. Does the Commission's prison and sentencing impact assessment include fiscal or other resource findings?

    The Commission's prison and sentencing impact assessment does not include fiscal or other resource findings. The Commission limits its analysis to estimated changes in average sentence and in hypothetical prison beds. The Commission cannot assess the cost of housing federal inmates, because housing costs are linked to the security status of the prison facility and the Commission has no ability to estimate an offender's security level. Nor can the Commission determine if there are currently available beds in prisons, at what point beds would need to be added, or the type of institution that would need to be constructed (e.g., minimum or maximum security) to increase inmate capacity. Consequently, the Commission's estimates do not include fiscal or other resource findings. When prison and sentencing impacts are requested by the Congressional Budget Office (CBO), the CBO will use the Commission's analysis to develop a cost/savings estimate along with additional information provided by the Bureau of Prisons.

  8. How far back does the Commission's prison and sentencing impact assessment web page go?

    The Commission's prison and sentencing assessment web page was launched on July 3, 2012, and contains information starting with analyses completed during the FY2012 amendment cycle. Going forward, new prison and sentencing impact assessments completed for legislation or federal sentencing guideline changes will be posted on the website.

  9. Can I get access to prison and sentencing impact assessments not included on the Commission's web page?

    Everything that the Commission has made publicly available is included on the Commission's website. Prison and sentencing impact analyses not included on the website are not available to the public. Analyses may not be made public for a number of reasons, including because the analysis was requested directly by a member of Congress, or because the analysis was performed for the Commission during the deliberation phase of proposed guideline changes and those changes were not promulgated by the Commission's final vote.