United States v. Marvin Leo Beasley

809 F.2d 1273, 22 Fed. R. Serv. 507, 1987 U.S. App. LEXIS 971

• Court: Court of Appeals for the Seventh Circuit
• Decided: January 9, 1987
• Docket: 86-1605
• Status: Published

Opinion

EASTERBROOK, Circuit Judge.

Meese, Inc., of Madison, Indiana, hired as a consultant Marvin Leo Beasley, whose Ph.D. is in chemistry. Beasley had some 75 scientific publications to his credit, including one — M.L. Beasley & R.L. Collins, Water-Degradable Polymers for Controlled Release of Herbicides and Other Agents, 169 Science 769 (Aug. 21, 1970)— that claimed a substantial scientific advance in the application of herbicides. Other publications, such as W.O. Milligan, M.L. Beasley, M.H. Lloyd & R.G. Haire, Crystalline Americium Trihydroxide, 24 Acta Cristallographica 979 (1968), and K. Maer, M.L. Beasley, R.L. Collins & W.O. Milligan, The Structure of the Titanium-Iron Cyanide Complexes, 90 J.Am. Chemical Soc. 3201 (1968), demonstrated Beasley’s accomplishments in his field. His new employer wanted to take precautions against the day it might lose access to Beasley’s skills and paid for a $1 million life insurance policy, with itself as beneficiary.

To obtain the policy Beasley had to submit the results of a physical examination. Warren Rucker, a physician who was also the Mayor of Madison, administered the examination. Beasley took the occasion to explain to Rucker one of his theories: that administering large doses of tranquilizers and analgesics to vegetables would help them deal with stress better and absorb nutrients more quickly, increasing their rate of growth. Beasley needed only the drugs to test this thesis. Dr. Rucker decided to help Beasley conduct his experiments and wrote out many prescriptions. Most were in Beasley’s name, with one each in the names F.E. Brooks and Marilyn Pierce, who Beasley said were his assistants and would need access to drugs while Beasley was out of town. The amounts of drugs prescribed, and the amounts Beasley obtained from three pharmacies between August 1980 and January 1981, are:

Drug_ Prescribed Dispensed

Dilaudid 4 mg 13,970. 7,470.

Dilaudid 1/24 gr 37. 37.

Codeine 1 gr 800. 800.

Morphine lk gr 1.300. 300.

Morphine Sulfate 15mg/ec 240. cc 240. cc

Percodan 1,100 600.

Demerol 100 mg 1.300. 300.

Preludin 75 mg 1,900. 1,000.

Desoxyn 15 mg 1,850. 850.

Desoxyn 5 mg 34. 34.

Desoxyn 2.5 mg 104. 104.

Quaalude 300 mg 800. 200.

Parest 400 mg 1,600. 800.

Tuinal 3 gr 100. 100.

Valium 10 mg 3.700. 3,000.

Librium 25 mg 700. 1,000.

Temíate 75 mg 100. 100.

Ionamin 30 mg 600. 530'.

Meprobamate 400 mg 800. 300.

Seconal 1.5 gr 600. 100.

Tetracycline 1,480. 1,280.

Premarin 2,100. 2,100.

V-Cillin 500. 300.

Librax 700. 700.

Penicillin 1.700. 1,700.

Terramycin 250. 250.

Amoxicillin 900. 100.

Lasix 300. O.

Vitamin B12 3,000. cc 1,650. cc

Estrogen 300. O.

Azo Gantrisin 500. 500.

Ampicillin 2,100. 1,400.

Nicotinic Acid 100. 100.

Procaine 8. 8.

Pontocaine 60. cc 60. cc

All of the drugs in the first group are controlled substances — either narcotics or. treated in the same way as narcotics.

Beasley says the vegetables took their medicine. The U.S. Attorney believes that the drugs were sold on the black market and turned at least one person into a vegetable. An indictment charged Beasley with seven counts of obtaining Dilaudid (hydromorphone HCL), a Schedule II controlled substance, with intent to distribute, in violation of 21 U.S.C. § 841(a)(1), and two counts of attempting to obtain Dilaudid by misrepresenting the name of the person to appear on the prescription, in violation of 21 U.S.C. §§ 843(a)(3) and 846. Beasley was convicted on all nine counts by a jury; the judge sentenced him to nine concurrent seven-year terms and fined him $1,000. Dr. Rucker and the three pharmacists who made all this possible were not charged; the U.S. Attorney believes they were simply credulous. But see 21 C.F.R. § 1306.-02(f) (defining a prescription as an order to dispense drugs “to or for an ultimate user”), and § 1306.04(a) (no prescription is lawful unless issued “for a legitimate medical purpose by an individual practitioner acting in the usual course of his medical practice”) — both of which suggest that even if Dr. Rucker believed every word Beasley told him (and the pharmacists believed Rucker, who relayed the tale), all knew or should have known that they lacked authority to distribute these drugs to Beasley. But there may be other matters of which we are unaware, and our function is to examine the claims of those who were convicted rather than to inquire into the position of those not charged.

Beasley mounts a frivolous attack on the two § 843(a)(3) convictions, contending that the evidence is insufficient because the prosecutor did not prove that Brooks and Pierce had not authorized their names to be used. The prosecutor correctly replies that such proof is unnecessary. The government showed that Beasley lied to Dr. Rucker in the course of obtaining these prescriptions — both about his relationship to Brooks and Pierce and about the use of the drugs. This made Beasley’s procurement of the prescriptions a knowing acquisition of controlled substances by fraud, whether or not Brooks and Pierce consented. See United States v. Hill, 589 F.2d 1344 (8th Cir.), cert. denied, 442 U.S. 919, 99 S.Ct. 2843, 61 L.Ed.2d 287 (1979). The fraud was on Rucker, not on Brooks or Pierce.

There is a more substantial challenge to the government’s trial strategy, however. The prosecutor was worried about the jury’s reaction to Beasley’s academic credentials and success as a biochemist, which lent verisimilitude to what would otherwise be a preposterous excuse for acquiring mountains of drugs. The other problem was that the principal evidence that Beasley distributed the Dilaudid he acquired in Indiana would come from F.E. Brooks, a convicted drug dealer. The word of a felon versus the word of a biochemist is not the strongest case. So the prosecutor decided to introduce extensive evidence that Beasley acquired and distributed drugs between 1981 and 1984. This presents questions under Fed.R.Evid. 404(b), which provides:

Evidence of other crimes, wrongs, or acts is not admissible to prove the character of a person in order to show that he acted in conformity therewith. It may, however, be admissible for other purposes, such as proof of motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mistake or accident.

It also presents questions under Fed.R. Evid. 403, which provides that relevant evidence may be excluded “if its probative value is substantially outweighed by the danger of unfair prejudice”.

The case that Beasley acquired Dilaudid with intent to distribute comes from the size and irregular manner of his purchases coupled with the testimony of F.E. Brooks that over a ten-month period starting in the summer of 1980, Beasley sold Brooks large quantities of the drug. Brooks testified that he bought between 50 and 600 tablets at a time, at a price of approximately $33 per tablet. According to Brooks, some tablets were given to his girlfriend, his son, and his daughter Marilyn Pierce; the rest were sold to strangers for between $40 and $50 per tablet. Other evidence came from Rocky Terrell, one of Beasley’s assistants, who said that although he helped Beasley plant test plots of many kinds of vegetables, he never added any solution to the plants or observed Beasley do so.

Some evidence showing that Beasley may have committed crimes other than those charged came in without objection. The prosecutor introduced the list of drugs given above, showing that Beasley acquired many drugs other than Dilaudid. The remaining evidence in the case was met by objections and portrays Beasley as acquiring and distributing drugs between March 1981 (two months after the last acquisition charged in the indictment) and some time in 1984. Christy Terrell testified that in March or April 1981 Beasley gave her codeine at his farm in Oklahoma and showed her how to ingest codeine or dilute it for injection. Rocky Terrell testified that in December 1982 Beasley showed him how to fake pain so that a physician would prescribe Dilaudid. Between December 1982 and March 1983, Rocky Terrell, Carol Parks, and on occasion Beasley went “shopping for doctors” in Texas, Kansas, Oklahoma, and Arkansas. Rocky Terrell and Parks would visit physicians’ offices in the hope of obtaining prescriptions for Dilaudid. When they did, they would either sell the tablets or split them with Beasley. Rocky Terrell also related a conversation with Beasley in 1982, attributing to Beasley the statement that he had “some kind of deal going” with a physician to hold Dilaudid tablets off the black market until the price rose past $55.

Parks supported Rocky Terrell’s testimony, although she said that the “shopping for doctors” may have lasted until 1984. She testified that she used some of the Dilaudid and gave some to Beasley. Between February 1981 and June 1981, she maintained, she received Valium and codeine directly from Beasley (although she had asked him for Dilaudid).

Finally, Margaret Walraven gave an explanation for the absence from trial of Marilyn Pierce, Brooks’s daughter, in whose apartment the Beasley-Brooks sales of Dilaudid took place. According to Walraven, Pierce was a Dilaudid addict who could not testify because, shortly before the trial began, she had been committed to Central State Hospital, a mental hospital in Oklahoma, apparently suffering from the effects of the drug.

Although counsel for Beasley objected to this evidence at trial, the principal consideration came beforehand. Counsel asked the judge to prohibit the prosecutor from using the evidence about “shopping for doctors” and from offering evidence that would tend to show that Beasley used Parks, Pierce, and Walraven as prostitutes. The judge stated that he would exclude any evidence that Walraven saw Pierce “shooting up Dilaudid outside the presence of Dr. Beasley.” The prosecutor promised to keep from the jury any evidence tending to show prostitution but argued that evidence tending to show that Beasley dealt in any of the kinds of drugs for which Dr. Rucker had written prescriptions should be admitted to show a “pattern”. The prosecutor stated: “with regard to the testimony of Rocky Terrell and Carol Parks, the government believes it is pattern evidence and it is admissible because it shows a pattern that started in our time [i.e., October 1980 to January 1981, the period covered by the indictment] and continued through the next — picked up again in the fall of 1982 through the spring of 1983”. The judge responded: “Well, that seems to me to be admissible pattern evidence.” After counsel and the court dealt with questions concerning the potential testimony of a physician who may have helped Beasley acquire drugs in Oklahoma, 1 the colloquy continued:

MR. MOLLOY [Beasley’s lawyer]: We have still got the question about these prior bad acts. Did the Court rule on that?

THE COURT: I thought I ruled.

MR. MOLLOY: I was afraid you had, too.

THE COURT: These other prior acts of running around the country.

MS. LOPOSSA [the prosecutor]: Actually they are subsequent acts, your Hon- or.

THE COURT: Whatever they are, subsequent acts, I think they are especially close in time to be admissible under Rule 404(b) which does not make any distinction as to whether the other acts are prior or subsequent to the time or times set out in the indictment.

The decision to admit the evidence rested on a belief that the bad acts showed a "pattern” of crimes “especially close in time”. Come the trial, however, the judge repeatedly told the jury that the sole purpose of admitting the evidence was to show Beasley’s “intent”. The judge tried to limit the jury’s consideration in cautionary instructions as the evidence was admitted, and one of the instructions in the charge repeated this limitation.

The prosecutor tries to defend the decision on appeal by resurrecting the argument about “pattern”. “Pattern” is missing not only in the instructions to the jury but also in Rule 404(b)’s list of permissible uses of bad act evidence. “Pattern” usually is a shorthand for a series of acts that collectively identify the offender — the ten bank robberies by a gang disguised by red polka dot bandannas, the series of counterfeit bills made by an engraver who never gets the Great Seal quite right, and so on. The pattern serves as the signature that enables the jury to determine that this offense, too, was committed by the defendant. See the summaries in McCormick on Evidence 449 (Cleary ed. 1972), and 2 J. Weinstein & M. Berger, Weinstein’s Evidence 11404[16] (1986 rev.). This use of pattern to show identity, or sometimes the extent and membership of a conspiracy, is the usual one in this circuit, as in others. E.g., United States v. Grabiec, 563 F.2d 313, 318 (7th Cir.1977); United States v. Krohn, 560 F.2d 293, 296 (7th Cir.), cert. denied, 434 U.S. 895, 98 S.Ct. 275, 54 L.Ed.2d 182 (1977); United States v. McCord, 509 F.2d 891, 895 (7th Cir.), cert. denied, 423 U.S. 833, 96 S.Ct. 56, 46 L.Ed.2d 51 (1975).

When the similarity of the acts is used to identify the culprit, the proximity of the acts in time is important. Given enough time, similar crimes will be committed by other people, and the value of the other acts as an earmark is diminished. Thus many cases use language emphasizing proximity. But the many bad acts of which Beasley was accused during trial were not similar in the sense of demonstrating a modus operandi. None of the other bad acts involved duping a physician into making drugs available for “experiments.” The episodes of “shopping for doctors” were similar to each other, but not to the crimes of which Beasley stood accused; Beasley’s provision of Valium and codeine to his associates is not like the fraudulent acquisition of Dilaudid. See United States v. Shackleford, 738 F.2d 776, 783 (7th Cir.1984) (evidence that does not show a “unique design or plan” or demonstrate “distinctive features” of the conduct is not admissible to show common plan). We therefore bypass the dispute that has occupied the parties about whether two or three years is “too remote” for the latter acts to be proximate in time. Compare United States v. Jimenez, 613 F.2d 1373, 1376 (5th Cir.1980) (one year after the crime charged is too long), with United States v. DeCastris, 798 F.2d 261, 265 (7th Cir.1986) (ten years before the crime charged is not too long). Questions about “how long is too long” do not have uniform answers; the answers depend on the theory that makes the evidence admissible. Here the acts are so dissimilar that their timing becomes unimportant.

Some language in the government’s brief suggests that any commission of similar crimes is the sort of “pattern” that permits the evidence to come in. On this reasoning, one drug offense may be used to prove another, although a bank robbery could not be admitted in a drug prosecution. The prosecutor admits as limits the factors enumerated in cases such as United States v. Draiman, 784 F.2d 248, 254 (7th Cir.1986): the bad act evidence must be clear and convincing, must show a similar act close enough in time, must have value that out weighs the risk of unfair prejudice (the Rule 403 standard), and must be used to show something other than the defendant’s propensity to commit similar crimes (the Rule 404(b) requirement). Yet in the prosecutor’s view, to show similarity is to negate prejudice and the tendency of the evidence to condemn by besmirching character. Not so. The requirements are distinct. A rule that a judge may admit all evidence that the defendant committed crimes of similar varieties produces the gravest risk of offending the central prohibition of Rule 404(b): “Evidence of other crimes, wrongs, or acts is not admissible to prove the character of a person in order to show that he acted in conformity therewith.” The inference from “pattern” by itself is exactly the forbidden inference that one who violated the drug laws on one occasion must have violated them on the occasion charged in the indictment. Unless something more than a pattern and temporal proximity is required, the fundamental rule is gone. This is why “pattern” is not listed in Rule 404(b) as an exception. Patterns of acts may show identity, intent, plan, absence of mistake, or one of the other listed grounds, but a pattern is not itself a reason to admit the evidence.

This brings into focus the instructions limiting the use of the evidence to proof of “intent”. Patterns of similar acts may show intent, just as they may show identity. See, e.g., DeCastris, 798 F.2d at 265; United States v. Stump, 735 F.2d 273, 275 (7th Cir.), cert. denied, 469 U.S. 864, 105 S.Ct. 203, 83 L.Ed.2d 134 (1984); United States v. Flick, 516 F.2d 489, 495 (7th Cir.), cert. denied, 423 U.S. 931, 96 S.Ct. 282, 46 L.Ed.2d 260 (1975). That the other bad acts came after the crimes charged in the indictment does not preclude their use to show intent. United States v. Peskin, 527 F.2d 71, 84 (7th Cir.1975), cert. denied, 429 U.S. 818, 97 S.Ct. 63, 50 L.Ed.2d 79 (1976); United States v. Moschiano, 695 F.2d 236, 245 (7th Cir.1982), cert. denied, 464 U.S. 831, 104 S.Ct. 110, 78 L.Ed.2d 111 (1983); United States v. Serlin, 707 F.2d 953, 959 (7th Cir.1983).

Intent was an issue here. Beasley testified that he bought the drugs to conduct experiments, fed the drugs to his plants, and never distributed drugs to anyone. If this is true, he did not purchase with intent to distribute. The prosecutor claims that Beasley sold the Dilaudid to Brooks and perhaps others and did not conduct or intend to conduct experiments on the effects of narcotics on rutabagas or cauliflowers. A demonstration that in 1982 and 1983 Beasley, Rocky Terrell, and Carol Parks bilked other physicians into prescribing Dilaudid, which they resold, tends to show that Beasley acquired the Dilaudid in Indiana with the same intent. At the same time, the evidence inescapably creates a risk of the forbidden inference, that a person who violates the law at one time has a bad character and therefore violated the law at a different time. So although the episodes of “shopping for doctors” were relevant to show intent, they also had a potential for “unfair prejudice” — which the advisory committee’s note to Rule 403 describes as “an undue tendency to suggest decision on an improper basis, commonly, though not necessarily, an emotional one.” The sale of drugs brings emotions into play, especially when the evidence reveals that the drugs may have ruined Marilyn Pierce’s life.

When the same evidence has legitimate and forbidden uses, when the introduction is valuable yet dangerous, the district judge has great discretion. There are no bright line rules; it is easy to identify polar cases but impossible to draw a line of demarcation. Appellate courts can contribute only modestly to the making of the best decision case by case. The decision must be made on the scene, and once the imponderables have been weighed there is little to be gained from weighing them again on appellate scales. The balance would not be systematically better the second time around, and the costs of second-guessing include new trials that may be less accurate as events become more remote. Trial judges have a comparative advantage because they alone see all the evidence in context, and the judicial system as a whole takes advantage of the division of labor. See DeCastris, 798 F.2d at 265; cf. Anderson v. City of Bessemer City, 470 U.S. 564, 574-75, 105 S.Ct. 1504, 1512, 84 L.Ed.2d 518 (1985).

Yet although appellate courts cannot often tell whether it was best, all things considered, to let in a given piece of evidence, it may be possible to tell whether the district court and the parties took the right things into account. A flaw in the process is easier to detect than is a flaw in the result — and over the run of cases the consistent operation of process is more important, too. The principled and just functioning of the judicial system depends on careful observation of the rules that focus attention on the proper grounds of decision. The rules deal in probabilities. We cannot know whether admission of the bad act evidence against Beasley changed the outcome or produced an improper conviction; we can be confident that repeated, careless use of bad act evidence will increase the probability of such unhappy outcomes.

The objective is not to enforce “rules”; Rules 403 and 404(b) establish standards rather than rules. It is to ensure that standards not be applied as if they were rules, as if they established mechanical indicia (such as “One drug offense may be used as evidence to prove any other”). The list of exceptions in Rule 404(b), if mechanically applied, would overwhelm the central principle. Almost any bad act evidence simultaneously condemns by besmirching character and by showing one or more of “motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mistake or accident”, not to mention the “other purposes” of which this list is meant to be illustrative. We therefore repeat the theme of what McCormick on Evidence at 453 calls the “wiser opinions” — including one of our own, United States v. Phillips, 401 F.2d 301 (7th Cir.1968) — that there must be a principled exercise of discretion. The district judge must both identify the exception that applies to the evidence in question and evaluate whether the evidence, although relevant and within the exception, is sufficiently probative to make tolerable the risk that jurors will act on the basis of emotion or an inference via the blackening of the defendant’s character. Discretion, when exercised, will rarely be disturbed.

The record in this case does not show that the district judge took into account the power of this bad act evidence to impugn Beasley’s character. The pretrial hearing was perfunctory. The district judge did not ask the prosecutor to explain what she expected to show. The only two comments by the court are “that seems to be admissible pattern evidence” and “[the acts] are especially close in time to be admissible”. 2 We have already explained why “pattern” evidence as such is excludable unless it shows one of the listed exceptions; the district court appeared to act on the opposite view. We have also explained why temporal proximity — if two or three years could be called “especially close in time” — does not support use of the evidence. And the court did not demonstrate an effort to determine the likely effect of the evidence in poisoning Beasley’s character. The effect on Beasley’s character may well have been the dominant one, indeed the principal reason why the prosecutor wanted to use the evidence. If Beasley’s Ph.D. and scientific achievements might suggest to the jury that Beasley had a good character, the prosecutor wanted something that cut the other way.

If the power of the bad act evidence to show intent were obvious, and the danger of improper inferences low, we would not be concerned by the lack of balancing on the record. Judges need not explain the obvious, even briefly. We are not dealing, however, with transparently admissible evidence. Some seems almost transparently inadmissible. Evidence that Beasley gave Valium and codeine to two people in 1981 is not proof of the intent with which Beasley acquired Dilaudid — not unless any drug offense proves any other. Proof that Marilyn Pierce was addicted to Dilaudid and had to be committed to a mental hospital was gratuitous, much more likely to provoke an emotional reaction from the jurors than to help them understand Beasley’s intent. Doubtless it was important to tell the jurors that Pierce, named on one prescription and whose apartment was a locus of sales, could not be present. It would have sufficed, however, for the judge to tell the jury that “for reasons beyond the prosecutor’s control Marilyn Pierce cannot attend the trial” and to warn the jurors not to draw inferences for or against either party because of her absence. Or the court might have told the jury that Pierce was in a hospital and for medical reasons could not attend. The full reason for her absence — and the proof of the reason through Margaret Walraven, another Dilaudid addict — invited all the wrong kinds of inferences. Only the evidence that Beasley, Parks, and Rocky Terrell went “shopping for doctors” to acquire Dilaudid between 1982 and 1983 (or 1984) substantially implies intent to distribute in 1980, and even here the potential for an improper inference is strong enough to make its admission as part of the prosecutor’s affirmative case not a foregone conclusion. The thoughtful exercise of discretion would have mattered. (Beasley took the stand and denied selling drugs to anyone. The prosecutor does not seek to support the admission of any evidence as impeachment, and so we do not discuss that subject.)

It was certainly a mistake to admit the evidence about Pierce’s mental condition; it was probably a mistake to admit the evidence about Valium and codeine; the use of the “shopping for doctors” incidents was sufficiently problematic that a more discriminating treatment by the district court was called for.

With respect to the two convictions for fraudulently obtaining Dilaudid, the error was harmless. Intent to distribute was not in issue on these counts. The government proved beyond any doubt that Brooks and Pierce were not Beasley’s research assistants, as he had represented, and that the prescriptions were not designed to obtain drugs for their lawful use. Convictions on these counts were inevitable. See United States v. Hasting, 461 U.S. 499, 509, 103 S.Ct. 1974, 1980, 76 L.Ed.2d 96 (1983).

The error is not harmless with respect to the remaining counts, however. An error (other than a constitutional error) is not harmless if it “results in actual prejudice because it ‘had substantial and injurious effect or influence in determining the jury’s verdict.’ ” United States v. Lane, 106 S.Ct. 725, 732, 88 L.Ed.2d 814 (1986), quoting from Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 1253, 90 L.Ed. 1557 (1946). The extensive use of other-crime evidence to show intent weighed heavily against Beasley. It was an important component of the prosecutor’s case. Tellingly, the prosecutor does not argue that any error was harmless.

Although two of the nine counts were not affected by the admission of the bad act evidence, we vacate the sentences on these counts as well. This will permit the district judge to resentence Beasley on the two fraud counts with knowledge that these counts supply the sole justification for imprisonment rather than being tag-along concurrent sentences. See United States v. Manzella, 791 F.2d 1263, 1270 (7th Cir.1986). If there should be a second trial, and Beasley again be convicted on charges of possession with intent to distribute, the district court will be able to recon struct a package of sentences. Beasley should not be resentenced on the fraud convictions until the final disposition of all counts.

REVERSED IN PART, VACATED IN PART, AND REMANDED.

1 . Beasley had been convicted in an Oklahoma court of obtaining Dilaudid, codeine, and Quaalude in October 1980 by misrepresenting to a physician that they would be used in research and misrepresenting to a pharmacist that he had the necessary permission by the state to use the drugs in research. The conviction was reversed solely because the evidence did not show that the codeine had been diluted, and there had been no expert testimony about the contents of Quaalude tablets. Beasley v. State, 665 P.2d 852 (Okla.Crim.App.1983). The district judge apparently believed that the reversal of the conviction precluded use of the episode as a similar act showing intent. He did not give a reason, and we do not know of one.

2 . The government moved to supplement the record by correcting the transcript of the pretrial hearing to show that the colloquy included a statement by the prosecutor that "[o]ne of the criteria for admissibility ... is the government’s need for the evidence”, to which the judge replied: "Oh, so there is no evidence of distribution.” The district judge never acted on this motion. If the exchange occurred, it adds nothing. The prosecutor did not explain the nature of the government’s "need”, and the judge’s answer reflects a misunderstanding of what the prosecutor expected to show about distribution. The government had direct evidence from F.E. Brooks and inferential evidence from the size and irregular procedure of the purchases. The prosecutor did not attempt to explain to the judge what evidence would be forthcoming, so the mistake is understandable.