Defendant Phillip Crockett disputes the quantity of cocaine for which he was held responsible by the district court when it determined his sentence for conspiracy to distribute cocaine. Crockett pleaded guilty to conspiring with Kenneth Gullo, David New and others over a period from May 1991 through January 1992. Crockett was the supplier of drugs to Gullo, New and the others, who traveled from Springfield to Chicago to purchase drugs from him and then returned to Springfield to resell the cocaine.

Crockett argues on appeal that the evidence upon which the sentencing judge relied to determine drug quantity was not sufficiently reliable. He also contends that the sentencing court improperly took account of a February transaction which was outside the scope of the charged conspiracy and was not properly categorized as relevant conduct under the sentencing guidelines.

While we agree that the district court's stated rationale for the inclusion of the February transaction is unclear and may be erroneous, we believe that any errors here are harmless and we affirm Crockett's sentence. With the exception of an apparent arithmetical mistake, the district court's determination of the quantity of cocaine involved in the charged conspiracy was clearly supported by the evidence. When the computational error is corrected, Crockett's base offense level is the same whether or not the February transaction is included in the quantity calculation. For this as well as for other reasons, any errors are harmless.

I. Reliability of the Evidence of Drug Quantity

Kevin Gullo and David New ran a cocaine-selling operation in Springfield, Illinois for a number of years. In January 1991, Gullo became acquainted with Michael Harris while he and Harris were in the Sangamon County Jail. After they were both released, Harris introduced Gullo to Crockett and Crockett began supplying cocaine to Gullo's operation. From May 1991 until July 1991 Gullo made frequent trips (sometimes as often as three times in one week) to Chicago to purchase cocaine from Crockett. In August 1991, New began to make these trips on behalf of the illegal enterprise. New and Gullo were at times accompanied on these trips by other individuals, including Tina Rakestraw, Melody Bernhardt and Walter Martin.

In calculating the drug quantity, the district court relied heavily on the presentence investigation report. That report was supported by the grand jury testimony of Gullo, New, Bernhardt, Martin and Rakestraw; by documentary evidence of telephone calls between Crockett and Gullo; and by the sentencing hearing testimony of Bernhardt and of law enforcement officers Reed and Jones. Most of the attributed amount was directly related to the charged conspiracy. Crockett was directly involved in each of the transactions which was counted. However, one transaction took place in February, after the charged conspiracy ended with the arrest of Crockett's co-conspirators. This transaction was a controlled buy arranged by New, who had begun to cooperate with the authorities. The relevant conduct guideline, USSG § 1B1.3, provided the basis for inclusion of this transaction in the sentencing calculation.

At the sentencing hearing, Crockett contested the quantity attributed to him. As proof he offered his own testimony and pointed to various perceived inconsistencies and weaknesses in the government's proof. The district court found Crockett's testimony as to drug quantities incredible, and we defer to that determination. We therefore turn to Crockett's critique of the evidence on which the district court relied.

The most important evidence presented by the government was the testimony of Crockett's co-conspirators, most of which was in the form of transcripts of the grand jury proceedings. This testimony paints a picture of regular, ongoing transactions from May 1991 until January 1992. In May, June and July, Gullo made the pickups from Crockett, initially in the company of Harris and later either alone or accompanied by one of the other members of the conspiracy. Beginning in about August 1991, New took over primary responsibility for the purchases, traveling first with Gullo and then, after he and Crockett became familiar with each another, alone or with one of the others. The quantities were initially small (two to four ounces) and the trips frequent (two or three times a week). Eventually the quantities were larger (a kilogram or more) and the trips more widely spaced (every week or ten days).

This basic picture was consistently presented by Gullo, New, Rakestraw and Bernhardt. Primarily on the basis of their testimony, the presentence report recommended that Crockett be held accountable for 15.0 kilograms of cocaine with respect to the charged conspiracy. The sentencing judge reduced this amount slightly based on Crockett's argument that the presentence report double-counted certain transactions, leaving a computed total of 14.9 kilograms for the period from May 1991 to January 1992. The February transaction involved just slightly over 2 kilograms, which were counted as relevant conduct, bringing the total quantity attributed to Crockett at sentencing to just under seventeen kilograms.

The sentencing judge first held Crockett accountable for 680.4 grams of cocaine for transactions during the period from May 1991 through July 1991. Before the grand jury, Gullo testified that, beginning in about May of 1991, he made three or four trips with Harris over a month-and-a-half period. He testified that at that time he was buying amounts ranging from two ounces to a quarter of a kilo. Gullo further testified that after that time he continued making trips, unaccompanied by Harris. The frequency of these later trips depended on the amount purchased. As Gullo testified, "If I made a trip and got a quarter key, I might go once or twice a week. If I would go with a half key or a key, I would go once a week." Gullo G.J. Tr. at 18-19.

Tina Rakestraw testified that she accompanied Gullo on the trips to Chicago after Harris ceased going and before New came into the picture in August. She testified that the initial quantities purchased were about eight to ten ounces (about a quarter of a kilogram), but that by the summer of 1991 the group was purchasing fifteen to twenty ounces (about half a kilogram) every two or three days.

The presentence report, which the sentencing judge adopted with one exception, estimated the amounts purchased during the period from May 1991 to July 1991 at two ounces per week for twelve weeks. Thus the total amount attributed for this period was a little over half a kilogram. Gullo was making the bulk of the runs during this period. Given the testimony of Gullo and Rakestraw that the amounts purchased ranged up to at least half a kilogram per week, this is an extremely conservative estimate.

David New began making the pickups from Crockett no later than August of 1991, according to the testimony of Rakestraw, Gullo and New. At the time New first became involved, he testified that the purchases were in amounts ranging from a quarter to a half kilogram. This testimony was entirely consistent with Rakestraw's statement that by the summer of 1991 the individual purchases had reached half-kilogram proportions. New testified that, toward the end, the quantities increased and that the largest purchase was for one and a half kilograms. This testimony was corroborated by Gullo. New also testified that the frequency of the trips ranged from three times a week, when the quantities were smaller, down to twice one week and once the next, when the quantities were larger.

In order to calculate the quantities of cocaine handled by the conspiracy in the months from August to December of 1991, the presentence report assumed that the trips were made twice one week and once the next over the entire period. On this basis there were about thirty trips during this period. Further, Gullo testified that, on at least five occasions after October 1991, the group purchased a kilogram of cocaine with money which it obtained in part from another co-conspirator, Jimmy Withers. Tina Rakestraw also testified that during December she traveled with New to buy one and a half kilograms plus an extra nine or ten ounces. The presentence report held Crockett accountable for about 6.5 kilograms as a result of these six specified trips.1 There was a small error in arithmetic in this calculation. The presentence report listed an item which was described as "1 trip for 1 1/2 kilograms." Actually, however, in adding up a number of purchases, this item was entered as 1,250.0 grams. This entry was apparently in error and should have been 1,500.0 grams. If this correction is made, the total quantity for the six particular trips should have been about 6.75 kilograms.

In order to estimate a quantity for the twenty-four remaining trips, the report assumed, quite conservatively, that each trip involved a purchase of a quarter kilogram. The total estimated from reports of the frequency of trips during this period was thus about 12.5 kilograms. Correcting for the arithmetic error that we have noted, the quantity involved in the transactions from August to December 1991, should have been about 12.75 kilograms.

In addition to these noted quantities of drugs, Crockett was held responsible for a purchase of about .7 kilograms, reported by Martin's testimony to have been made in January, and for one kilogram based on a wrapper discovered in the garbage at Gullo's residence in January.

The total quantity attributed to Crockett for the period from May 1991 to December 1991 was thus 14.9 kilograms. As noted, the correction of an error of arithmetic brings this total to approximately 15.1 kilograms. Crockett's responsibility for this quantity is amply supported by sufficiently reliable evidence. As noted, it is plain from the evidence presented that 15.1 kilograms is a very conservative estimate of the total quantity of drugs involved in the Springfield-Chicago conspiracy. Further, the sentencing judge did not simply accept the numbers in the presentence report, but agreed with the defendant's argument that some double-counting might have occurred, and reduced the quantity accordingly. The 15.1 kilogram total corresponds to less than half a kilogram per week when averaged over the entire life of the charged conspiracy.

The grand jury testimony of Rakestraw, Gullo and New could certainly have supported a higher estimate since the presentence report calculations did not take into account the correlation between more frequent trips and smaller purchases. When that correlation is taken into account, the testimony suggests that, beginning in August 1991, the weekly quantities were at least three-quarters of a kilogram. Similarly, if the reported correlation between quantity and frequency during June and July were taken into account, the testimony could support a finding that there were transactions totaling at least one-fourth kilogram per week. If the sentencing judge had used these quite reasonable estimates, the total quantity attributable to Crockett for the charged conspiracy would have been nearly 27 kilograms.

Crockett argues that the quantity determinations are unreliable since they are based on the testimony of admitted drug users; since they are based on hearsay; and since they are based on conflicting testimony. Heavy drug use is certainly a factor to be weighed in ascertaining the reliability of witnesses. However, that factor can be weighed against others, such as the consistency of the testimony of the several participants in the scheme (Gullo, New, Rakestraw and Bernhardt), in determining whether to credit the testimony. Credibility of witnesses is for the district court, which had a plethora of evidence showing the large scale of this conspiracy. That evidence, contrary to Crockett's assertions, was quite consistent and was also consistent with the frequent telephone traffic. There was no clear error in relying on the testimony of Crockett's co-conspirators.2

Crockett also complains that most of the evidence of drug quantity was hearsay, having its source either in the grand jury testimony of his co-conspirators or in the sentencing hearing testimony of law enforcement agents relating the statements of co-conspirators to the police. Hearsay, however, is admissible in sentencing proceedings, as long as it is sufficiently reliable. Grand jury testimony, having been made under oath in a formal proceeding, is a particularly reliable form of hearsay. Here the grand jury testimony of the various witnesses was consistent with the testimony offered by Melody Bernhardt at the sentencing hearing. Her testimony was given in connection with the phone records, which corroborated the frequent interactions between Crockett and the Gullo group and with the testimony of the officers about the statements made by various co-conspirators to the police. In this instance, the grand jury testimony was clearly reliable enough to form a basis for making the quantity determination. The hearsay testimony of the officers corroborated the grand jury testimony.

Finally, Crockett argues that there are inconsistencies between the various statements of his co-conspirators. For example, in his brief Crockett characterizes as an "inconsistency" the fact that, although the members of the group often took the train from Springfield to Chicago, train tickets for only three dates were uncovered in the search of Gullo's apartment. The fact that more train tickets were not discovered is not inconsistent with there having been more frequent trips to Springfield. First, many of the trips were made by automobile. Second, although Officer Reed testified that Gullo "kept a lot of papers ... didn't throw anything away[; n]ormal day-to-day," and "didn't clean the house a lot," there is no indication that train tickets were saved in any systematic way. Sentencing Tr. at 59. (Indeed, one can hardly imagine that they would have been.) There is similarly no reason to expect that all train tickets would have even been brought to the house, rather than discarded somewhere along the way home. Thus, the fact that only a few tickets were found is not inconsistent with many more trips having been made. Crockett also makes much of the fact that the grand jury testimony indicated that the cocaine wrappers used by Crockett were of various colors, but the wrapper retrieved from Gullo's garbage was clear. Crockett argues that the court erred in finding that this clear wrapper came from cocaine which he supplied to Gullo. However, the wrapper was discovered by police only after Rakestraw told the police to check in the garbage for it. Officer Reed testified that Rakestraw told him that the wrapper came from Crockett. Reed also testified that New and Gullo said they had once received cocaine from Crockett in a clear wrapper. In addition, Bernhardt (who was Gullo's girlfriend and, later, wife) testified that, to her knowledge, Crockett was Gullo's only supplier and that she would have known if he had any other suppliers. This evidence is sufficient to prove, by the preponderance standard employed at sentencing, that the cocaine which had been packaged in this clear wrapper came from Crockett.

Crockett also argues that there are inconsistencies between the statements of his various co-conspirators. Upon close inspection, none of the purported inconsistencies are significant enough to seriously affect the quantity determination. For example, Crockett states in his brief that the presentence report, at paragraph 26, "had New purchasing drugs from Defendant in July 1991." Def. Br. at 12. In fact, paragraph 26 says that "[b]eginning in August, 1991, New, who began working for Gullo, was making the trips to Chicago." As another example, Crockett asserts that New's testimony at New's own sentencing hearing was inconsistent with his grand jury testimony. At the sentencing hearing, New stated that it was October 1991 (rather than August 1991) when he began making trips to Chicago. It is true that this testimony, given in August 1992, is inconsistent with the grand jury testimony, indicating August 1991, that New gave in January 1992. In any event, the August starting date is consistent with all of the other testimony. The sentencing judge did not clearly err in deciding to credit the much more abundant evidence of an August starting date for New's involvement.

In sum, the drug quantity estimate for the May 1991 to January 1992 period of the charged conspiracy was carefully and conservatively made. Indeed, it is likely that the presentence report underestimated the quantity actually involved in the string of transactions by as much as a factor of two. Thus, Crockett's challenge to the quantity determination for May 1991 to January 1992 fails. Our discovery of an arithmetic error, however, requires that the quantity estimate for this period be revised upward by .25 kilograms to a total of about 15.1 kilograms.

Crockett raises an entirely different, and more substantial, objection to the inclusion of the February transaction in the quantity determination. The February transaction resulted from the arrest of Gullo and New in January. After being arrested, New agreed to cooperate with the government by setting up a controlled buy from Crockett. New called Crockett and set up a meeting among Crockett, New and an undercover government agent. According to the agent, Crockett gave New 6.1 grams of cocaine as a sample prior to the meeting. At the meeting, the agent negotiated with Crockett to purchase two kilograms of cocaine for $48,000. The agent reported that Crockett agreed to the transaction but wanted the money up front. The transaction was never brought to completion.

At sentencing, two kilograms plus 6.1 grams were attributed to Crockett because of this February transaction. Crockett argues that this transaction may not be treated as relevant conduct for his conviction of a conspiracy which ended in January 1992. He argues that the postconspiracy transaction cannot qualify as conduct relevant to the conspiracy--which had terminated with the arrest of Gullo and New--and, in addition, that the district court failed to articulate the necessary findings that the transaction was relevant conduct. Indeed, Crockett contends that the district court relied upon an erroneous definition of relevant conduct in making its ruling.

The district court's discussion of this point was as follows:So you have a series of actions, jointly taken, let's say over a two-year span of time, with Mr. Crockett dealing with New and then with somebody else, and somebody else. Then, but carved right out of the center of it is a particular conspiracy timeframe as charged with New and Crockett and ... Gullo.

Now, as far as counts of the indictment, as far as the charge of conspiracy, the timeframe of those six months is right out of the center. But the sentencing guidelines, in my estimation, contemplates you go beyond that, before and after the conspiracy.

And even though not charged, maybe a new organized conspiracy, maybe changes of players, whatever it is, it is still related conduct dealing in the field of controlled substances and drugs, for the purpose of sentencing.

Sentencing Tr. at 135-36.

The determination whether a particular drug transaction is relevant conduct to the offense of conviction is guided by USSG § 1B1.3. This guideline provides that, for offenses for which "the offense level is determined largely on the basis of ... the quantity of a substance involved," USSG § 3D1.2(d), relevant conduct includes "all acts and omissions committed ... by the defendant," USSG § 1B1.3(a)(1)(A), that were "part of the same course of conduct or common scheme or plan as the offense of conviction." USSG § 1B1.3(a)(2). The commentary to § 1B1.3 further defines "common scheme or plan" and "same course of conduct" as:

(A) Common scheme or plan. For two or more offenses to constitute part of a common scheme or plan, they must be substantially connected to each other by at least one common factor, such as common victims, common accomplices, common purpose, or similar modus operandi.

...

(B) Same course of conduct. Offenses that do not qualify as part of a common scheme or plan may nonetheless qualify as part of the same course of conduct if they are sufficiently connected or related to each other as to warrant the conclusion that they are part of a single episode, spree, or ongoing series of offenses. Factors that are appropriate to the determination of whether offenses are sufficiently connected or related to each other to be considered as part of the same course of conduct include the degree of similarity of the offenses, the regularity (repetitions) of the offenses, and the time interval between the offenses....

As we have noted on prior occasions, the government's burden at sentencing is considerably lightened through the application of the preponderance of the evidence standard and the relaxed evidentiary rules which govern. However, because the relevant conduct rule is "not without limits" and "because its application so favors the government," we insist that courts be "scrupulous to ensure that the government has adhered to those limits." One of the ways in which we ensure that these limits are maintained is by requiring sentencing courts to "explicitly state and support, either at the sentencing hearing or (preferably) in a written statement of reasons, the finding that the unconvicted activities bore the necessary relation to the convicted offense." United States v. Beler, 20 F.3d 1428, 1431-32 (7th Cir.1994), citing United States v. Duarte, 950 F.2d 1255, 1263 (7th Cir.1991).

One important distinction to be made in the relevant conduct determination is "that section 1B1.3(a)(2) should not be applied to offenses that are of the same kind, but not encompassed in the same course of conduct or plan as the convicted offenses." Beler, 20 F.3d at 1432, quoting United States v. White, 888 F.2d 490, 500 (7th Cir.1989). The mere fact that the defendant has engaged in other drug transactions is not sufficient to justify treating those transactions as "relevant conduct" for sentencing purposes. To ensure that this distinction is maintained, we require that district courts make specific findings that offenses to be included as relevant conduct for sentencing purposes are part either of the same course of conduct as the charged offense or of a common scheme or plan including the charged offense.

The sentencing judge in this instance was less than precise in his statement supporting the finding that the February transaction constituted relevant conduct in that it was part of the "same course of conduct or common scheme or plan." The court's statement that the February transaction was "related conduct dealing in the field of controlled substances and drugs" could be construed as relying on the same sort of reasoning that was rejected in White. It is not, of course, enough that the conduct in question merely involved drugs. Here, however, other parts of the statement seem to refer (albeit indirectly) to closeness of time and similarity of participants as grounds for including the February conduct. These comments would be germane to the critical question whether the February transaction was part of the same course of conduct or common scheme or plan as the charged conspiracy. And certainly the objective circumstances here are such that a district court would have no difficulty in invoking closeness of time and similarity of participants.

The determination that an uncharged drug transaction is to provide a basis for an increased sentence of imprisonment is, of course, a very significant matter. The rules delineating the circumstances under which such transactions may be included in the drug quantity computation must be carefully observed. To ensure that only relevant conduct is included in calculating the sentence, the government must carefully present the justification for the inclusion of any uncharged acts and the court must, in turn, set out its reasoning as clearly as possible. The conclusory statement that "whatever it is, it is still related conduct dealing in the field of controlled substances and drugs" is not adequate.

The deficiencies in the district court's statements, however, are harmless for three separate reasons: First, the available evidence can easily support a finding that the February transaction was part of a "common scheme or plan" with the Gullo-Crockett-New conspiracy. The February transaction involved a common accomplice--New; a common purpose--cocaine sale; and a common modus operandi--meeting in a restaurant and then proceeding to another location to transfer the drugs.4 The transaction took place in close temporal proximity to the sales included in the charged conspiracy. Second, Crockett's base offense level would not, in any event, be affected by the exclusion of the 2 kilograms and 6.1 grams of cocaine associated with this transaction. The final quantity determination at sentencing was 16.87 kilograms. As already discussed, this result was infected with an arithmetic mistake. When that mistake is corrected, the quantity of drugs attributed to Crockett by the district court for the charged conspiracy alone is approximately 15.1 kilograms. This quantity is sufficient to yield the same base offense level of 34 upon which Crockett's sentence was based. Thus, the sentencing court's questionable statement of reasons for including the February transaction as relevant conduct had no effect on Crockett's sentence, since, with or without that transaction, the cocaine attributed to Crockett would have exceeded 15 kilograms.

In conclusion, we affirm Crockett's sentence since it was supported by reliable evidence that the charged conspiracy involved more than 15 kilograms of cocaine. As we have noted, the district court's statement of reasons for including the cocaine sold in February was inadequate. However, whatever deficiencies existed in this determination of relevant conduct, the errors are harmless, and do not affect Crockett's sentence. His sentence is therefore affirmed.

AFFIRMED.