About the Data
Understanding the Terminology
Which Agencies Use
Federal enforcement agencies undertake a variety of activities which together make up the federal criminal enforcement system. In order to track and manage these activities, specialized terminology is often adopted to classify these activities into meaningful categories, and counting rules devised to measure their volume. With office automation and the adoption of computerized record keeping systems, these categories and rules must be formalized and are often elaborate.
Understanding the specific definitions of categories used, along with an agency's counting rules, are important in understanding the "numbers" an agency comes up with to describe its activities. When more than one agency or office within an agency are involved in the same activities, different classification systems or differences in counting rules can lead to differences in the figures each office publishes about identical sets of events. This can be very confusing. But it need not mean that one or the other of these figures is wrong. It is possible to have two different numbers if each office has chosen to classify and count events somewhat differently.
Added confusion can occur because of the complexity of the criminal justice system itself. Often real distinctions among different, although related events, may not be appreciated by someone who is not familiar with how the system operates. The uninitiated may encounter two sets of figures which they assume to be referring to the same thing but which are in fact referring to different events or different distinct stages in the process. In this situation, the cause of the confusion comes from an incomplete understanding of a complex process.
To reduce such problems, anyone who uses agency figures, or works with the underlying data, must work hard to understand the specialized terminology used by that agency to classify and count events. For a detailed understanding one must often turn to the agency's technical documentation about its data systems. This page provides a brief guide to basic terminology issues, highlighting important distinctions and common choices databases face in classifying and counting events in the federal criminal enforcement system. Specific reference is made to definitions used in the TRAC's federal prosecutors database which is a core data source used on this website.
Investigations, Referrals and Matters. Investigative agencies such as the IRS conduct criminal investigations of suspected wrong-doing. Those investigations which result in evidence the agency feels adequate for criminal prosecution are referred to the Justice Department with a recommendation that the individuals involved be criminally prosecuted. Federal prosecutors in the Justice Department then decide whether or not they will prosecute, and if they decide to prosecute, what specific charges they will bring.
Federal prosecutors classify referrals they receive into two general categories. Referrals may be declined immediately ["immediate declinations"] or considered further by the office for possible criminal prosecution. These latter are officially called "matters." Officially, a referral becomes a "matter" when an assistant U.S. Attorney spends an hour or longer considering or working on the referral. At that point a file is opened and matters are tracked more elaborately through to final disposition.
Just because an assistant United States attorney treats the referral as a "matter," does not mean that a prosecution ultimately will take place. A large number of referrals accepted as "matters" are ultimately declined ["declinations"]. In very rare instances, the federal prosecutor seeks an indictment and a grand jury refuses to indict. Otherwise the referral moves forward and becomes a "case" -- an actual prosecution.
Aspects of the recording of these events complicate this basic outline. First, there is some subjectivity in what gets recorded as a referral, particularly when it is immediately declined. Further, the division of referrals into immediate declinations and matters in practice is also a bit subjective. U.S. Attorney accounting practices appear to vary -- both among district offices and over time -- regarding which referrals are recorded and, when recorded, whether they are classified as an immediate declination or a matter. For whatever reason, the number of immediate declinations recorded by the U.S. Attorneys has declined sharply since 1980. Increasingly, the referrals that are recorded at all are being treated as matters, even when they are subsequently declined. Some caution is therefore warranted when comparing referral counts.
A further complication is that federal prosecutors often do not employ their own terminology consistently when discussing their activities. This is a problem even in the agency's official publications. In contrast to how agency personnel are instructed to record events in their own databases as discussed above, the official Statistical Reports for United States Attorneys' Offices sometimes use the terms "referrals" and "matters" as if they are interchangeable, and sometimes draw a distinction. [See, for example, p. 4 of their 1995 report which at several places refers to "matters" and "referrals" as the same thing: "The United States Attorneys receive most of their criminal referrals, or "matters," from federal investigative agencies..." and "Matters received includes immediate declinations in addition to later declinations and files initiated in any court."] Thus, TRAC finds it is necessary to reconstruct the agency's published statistics using the agency's own databases to determine just what the agency is counting and when, to attempt to ensure that more consistent labeling is used.
Prosecutions versus Declinations. Federal prosecutor data systems rather than distinguishing between referrals on the basis of whether or not they are prosecuted, classify disposition reasons on the basis of whether or not prosecution was declined. There is a small class of referrals which are not prosecuted (or even brought before a Grand Jury), but in the narrow sense federal prosecutors use, are not declined either. (This residual category includes transfers among districts. Because separate data systems are maintained by individual districts, it is not possible in the information provided to match up transfers. ) TRAC classifies dispositions into "prosecuted" versus "not prosecuted." The latter category is used synonymously with "declination."
Federal prosecutors also follow a somewhat confusing practice in what gets recorded as a "case." If the U.S. Attorney office decides to prosecute through regular court action, the "matter" becomes a "case." However, prosecutions in Magistrate Court are handled somewhat differently in their tracking systems and retain a "matter" designation. "Proceedings that are conducted before a Magistrate and which are assigned Magistrate Court numbers, such as misdemeanor information, are categorized as matters, not cases, even though they involve a charging instrument and may involve court time." [Statistical Report, United States Attorneys' Offices, Fiscal Year 1992, p. 4.]
Less information is recorded on "matters" than on "cases" so it is not usually possible to determine from the database whether a prosecution in Magistrate's Court has taken place until the prosecution is completed. (This is also an area where some inconsistencies in coding have been noted by TRAC -- see federal prosecutors.)
Again, however, federal prosecutors do not always employ their own terminology consistently when discussing their activities. Thus, for example, a table [Table 2A] from their 1995 published annual statistical report refers to "Disposition of Criminal Cases and Defendants in U.S. Magistrate Courts," even though agency databases classify these as "matters" not "cases." As noted above, TRAC reconstructs the agency's published statistics using the agency's own databases to determine just what the agency is counting and when to attempt to ensure that more consistent labeling is used.
Referral versus Disposition Counts. Because the date that a referral is received and the date that it is ultimately disposed of may occur in different fiscal years, the counts for the number of referrals received in a given year generally will differ from the counts for the number of referrals disposed of in that year. Many referrals received during a year will still be part of the active workload at the end of the year. Referrals disposed of during a given year will include referrals received in prior years. While some referrals will be received AND disposed of during the same year, many of the referrals that make up these two separate counts will in fact involve separate defendants and offenses. Because of this, it is not unusual for the number of dispositions to exceed the number of referrals in a given year. (See also discussion of "Dates Used to Classify Events" on this page.)
Thus, when using these numbers to calculate percents and rates, it is important that the appropriate counts be combined in the calculation. When examining the percent of referrals that result in conviction or prison sentences, the base for this calculation is the number of referrals DISPOSED of that year. Only on disposals, will such outcome information be known.
Prosecutions Filed versus Prosecutions Completed. Similar care must be used in examining percents and rates of criminal prosecution. Because the date a prosecution is filed and the date that the prosecution is completed may occur in different fiscal years, the counts for the number of prosecutions filed in a given year also generally differ from the number of prosecutions disposed of that year. ["Prosecutions disposed of" are the sum of the number of convictions plus the number not convicted.]
When examining convictions rates -- the percent of prosecutions resulting in a conviction, the base for this calculation is the number of prosecutions disposed of that year. Otherwise, outcome of the prosecution would not be known. However, the prosecution rate -- that is, the percentage of referrals on which a prosecution is filed -- compares the number of prosecutions filed during the year with the number disposed of without prosecution. [More specifically, the percent filed divides the number of prosecutions filed by the sum of the filings plus declinations and then translates this into percentage terms. The percent declined divides the number of declinations by the sum of the number of prosecutions filed plus declinations.]
Dates Used to Classify Events. Because of the natural delay between the occurrence of an event and its recording in the agency's database, two parallel date records are kept. The first is the actual date -- such as for when the referral is received, the case is filed, or the disposition occurs. The second is the recording date -- when the event was added to the computer system. The Justice Department uses the date the event was added to the computer system to classify whether the event occurred during a given fiscal year. This may result in events which took place prior to the fiscal year in question, being listed in the current year's file.
Counting with Single vs. Multiple Defendants. An individual referral or court case often involves multiple defendants. A separate record is maintained for each potential defendant for matters and cases. For immediate declinations, however, a separate record is included in the original Justice files only for each referral, although a field indicates the total number of suspects involved. To maintain consistency in definitions, TRAC counts each suspect recorded in an immediate declination to ensure that summary counts and rates are based upon the number of individuals involved for all referrals.
Classifying Prosecutions by the Underlying Charge(s). Most data systems classify enforcement activities by the underlying charge(s). When more than one charge is involved, the most serious or so-called lead charge is commonly used. Different data systems use different criteria for determining which of the charges to pick as the most serious or lead charge. The U.S. Courts and the U.S. Sentencing Commission use the charge with the maximum statutory penalty that can be applied; the U.S. Prisons classify by the most serious penalty in fact applied. The federal prosecutors use a more subjective rule: "the substantive statute which is the primary basis for the referral" where primary is left for determination to the individual assistant U.S. attorney who is assigned the case. Note also that the classification takes place at the time the referral is received. It is thus possible that the defendant may not ultimately be prosecuted on this charge.
Classifying Prosecutions by Investigatory Agency. Federal prosecutors classify referrals by lead investigatory agency. In more recent years, when the investigation was jointly carried out by the agency with a state or local task force that distinction is also noted. However, when several federal investigatory agencies are involved in the investigation, federal prosecutors must attribute the referral to a single so-called lead agency. Which agency is the lead agency is again left to the determination of the individual assistant U.S. attorney handling the case. Counts of referrals provided by investigatory agencies may therefore differ from the counts from federal prosecutors since each agency in a joint investigation may claim credit for the referral. Since 1998, optional fields have been provided where federal prosecutors now can record other agencies involved in a joint investigation.