Full opinion text
LYNCH, Circuit Judge. William Thurston, a vice president of Damon Clinical Testing Laboratories, Inc., appeals his conviction for conspiring to defraud the Medicare program of over five million dollars. The charged conspiracy involved the manipulation of physicians into ordering unnecessary ferritin blood tests for Medicare beneficiaries in violation of 18 U.S.C. § 371. The essence of the scheme charged was that Damon, through Thurston and others, bundled the ferritin blood test — previously ordered by doctors less than two percent of the time — into a panel of blood tests known as the LabScan, which was ordered thirty to forty percent of the time. When doctors or patients (instead of insurers) paid for the bundled LabScan, Damon provided the ferritin test for free, leading doctors to believe there was no extra charge for this test. Doctors were not told that, when Medicare paid for the bundled LabScan, Medicare was charged extra for the ferritin test. Indeed, both a letter and marketing materials indicated the added ferxitin test was “free”; that is, that there was no charge beyond the standard Lab-Scan charge. Those unnecessary ferritin tests were not free to Medicare. Damon charged Medicare roughly $21 per ferritin test on top of the approximately $24 charged for the LabScan. Nor were doctors told that the ferritin test could be ordered separately; the test requisition form did not offer that option. The physicians, then, were induced to order and to certify as medically necessary a large number of ferritin tests that were not medically necessary. The government’s theory was that Damon did this to offset Medicare’s 1988 reduction in reimbursement rates of sixteen percent, which was projected to cause Damon an estimated annual loss of $800,000 in revenues. In just one of Thur-ston’s labs, the orders for ferritin tests were expected to increase from approximately three hundred per month to roughly ten thousand per month. Thurston testified that he was innocent, and that he neither had knowledge of nor responsibility for key components of the conspiracy. The jury disagreed. Although the sentencing guidelines called for a sentence of sixty-three to seventy-eight months, the district court sentenced Thurston to only three months’ imprisonment. It did so by granting a downward departure to “correct” a perceived disparity between a five-year sentence of imprisonment for Thurston and the sentence of three years’ probation given to the company president, Joseph Isola, who had pled nolo contendere and assisted the government. Its second ground for departure was its sense that the good works Thurston did for his church and community were extraordinary. This sentence outraged the prosecutors, who appealed, arguing that the district court lacked the power to depart downward for any of the reasons it gave, and that, even if a departure were appropriate, the extent of the departure was excessive. The government also appeals the district court’s failure to impose a fíne, on the basis that the sentencing guidelines, if not the statute, mandated a fine. Finally, the prosecution argues that if some departure for good works was warranted, the district court was required to address an issue it avoided: the government’s request for an upward departure on the ground that Thurston had obstructed justice by committing perjury on the witness stand. Thurston also appeals, arguing that the conviction must be vacated because the prosecution was barred by the statute of limitations, because he was entitled to a jury instruction and to acquittal on the basis that he reasonably interpreted the law to mean he could rely on the physicians’ certifications of medical necessity, and because of other errors. In addition, Thurston argues that his sentence was too high because his base offense level calculation depended on an amount of loss, either actual or intended, that was unproven and excessive. Thurston also defends the downward departure. Several important issues are raised by these appeals. The government’s appeal requires us to address the effect of the new Prosecutorial Remedies and Tools Against the Exploitation of Children Today Act of 2003 (PROTECT Act), Pub.L. No. 108-21,117 Stat. 650, on the standard that the courts of appeals use to review downward departure decisions by district judges in sentencing, as well as the availability of a downward departure for a record of good works under U.S.S.G. § 5H1.11. Thurston’s appeal invites, inter alia, clarification of the defense doctrine concerning a defendant’s reasonable interpretation of the law; the issue of when a statute of limitations defense must be raised; the ramifications of a trial judge’s failure to respond to jury instructions proposed by counsel; the evidence needed to show an intended loss; and the question whether fines are mandatory. In the end we sustain the conviction but find that the sentence was in error. I. FACTS We state the facts as the jury could reasonably have found them, including a fair description of the defense evidence. A. Background Medicare provides certain medical services and care, including clinical laboratory testing services, to persons aged sixty-five and older and to persons with disabilities. At the time of the conspiracy, Medicare was administered by the Health Care Financing Administration (HCFA), a division of the U.S. Department of Health and Human Services. HCFA in turn contracted with private insurance companies (“carriers” and “intermediaries”) to handle claims for reimbursement to Medicare program beneficiaries. By law, Medicare only reimburses clinical laboratory services if those services were medically necessary for the treatment or diagnosis of a beneficiary’s illness or injury. 42 U.S.C. § 1395y(a)(l)(A) (2000); see also 42 C.F.R. § 424.10(a) (1988). Medicare did not generally reimburse screening tests. Medicare reimbursed one-hundred percent of the cost of necessary clinical blood tests. Damon was, at the time, a Massachusetts corporation that provided clinical laboratory testing services to physicians, hospitals, health maintenance organizations, and their patients nationwide. Approximately thirty percent of Damon’s revenues were from Medicare. Damon owned and operated a national system of clinical laboratories, and Damon was an approved Medicare provider. When Damon billed the Medicare program, it submitted to the carriers a HCFA 1500 form saying that it certified the lab tests were medically necessary. Physicians did not see those bills. Thurston sewed as Regional Vice President of Damon from 1987 to 1990, and was responsible during all or part of this time for Damon’s regional laboratories in New-bury Park, California; Phoenix; Chicago; and San Francisco. His office was at the Newbury Park lab; he traveled to the other labs and was in regular phone contact. Thurston was promoted to Senior Vice President of Operations in 1990. Thereafter, he relocated to the company headquarters in Needham, Massachusetts and supervised only the San Francisco lab. Another company purchased Damon in August 1993; Thurston later switched employers and moved to Utah. B. Theories of Prosecution and Defense The government’s theory was that the defendants tricked doctors into ordering medically unnecessary tests, for which Medicare paid. Damon added little-used tests to more popular panels of tests; it then tricked doctors by concealing that the panels could be ordered without the added tests and that Medicare was being charged for the tests. To conceal from doctors that Medicare was being charged, and to encourage doctors to order ferritin tests regardless of medical necessity, the defendants charged doctors and patients little or no extra fee for the added tests; provided literature saying that ferritin was provided free; otherwise failed to disclose to doctors the cost to Medicare of the expanded LabScan; and made it difficult to order the bundled test without ferritin. The defense theories were that there was no conspiracy and, if there were, that Thurston was not a knowing participant. The defense contended that Damon complied with existing Medicare regulations, which neither prohibited bundling nor imposed a requirement on the lab to disregard a physician’s certification that tests were medically necessary. It further argued that even if Thurston’s interpretation of the regulations was incorrect, it was objectively reasonable, and so Thurston lacked the required criminal intent. Thur-ston’s defense was also that any fraud was carried out by his subordinates, without his knowledge. He testified that he did not instruct the labs to add ferritin; did not authorize or condone any decision to forego a fee increase for doctors or private-pay patients on the expanded Lab-Scan; did not instruct subordinates to conceal that doctors could order a LabScan without ferritin; and had nothing to do with requisition forms or particular pieces of marketing literature. C. Evidence of Conspiracy to Defraud 1. Addition of Ferritin In late 1987, HCFA announced that effective April 1, 1988, Medicare would reduce by almost sixteen percent the fees paid to laboratories, including Damon, for providing clinical laboratory services to beneficiaries. If Damon maintained its existing practices and fee structure, then Damon would lose $800,000 in revenues during the first year alone, as Thurston knew. Of that amount, more than $500,000 of the annual losses would occur at Thurston’s four regional laboratories. There were corporate discussions, which included Thurston, about how to offset this loss. As a result, Damon added its ferritin test to the LabScan, a panel of more than a dozen blood chemistry tests performed by a single machine on one blood sample. The ferritin test was performed on separate equipment. Damon bundled ferritin with the LabScan from 1988 to at least mid 1993. The LabScan had been requested on at least thirty-five to forty percent of the orders submitted to Thur-ston’s regional laboratories. By contrast, doctors rarely ordered the ferritin test. The general manager of one of the laboratories Thurston oversaw estimated that only one to two percent of the orders for blood tests included a request for a ferritin test. Nothing in the medical literature at the time showed ferritin was necessary for all persons receiving LabScans. Indeed, it was not. A family practitioner testified that he needed ferritin less than ten percent of the time it was included as part of the LabScan. Similarly, an internist testified that he needed the ferritin test for very few of his Medicare patients. The plan to add ferritin to the LabScan was discussed at a general managers meeting in January 1988 and at a mid-year financial meeting for the western region in March 1988. Thurston attended both meetings. In late March, Thurston personally approved the addition of ferritin to the LabScans offered by the Newbury Park, San Francisco, and Phoenix labs. Thurston initially acquiesced when the Phoenix lab sought permission not to add ferritin; in August 1988, however, after Thurston saw the financial results from the addition of ferritin at his other labs, Thurston ordered the Phoenix lab to add the test to the LabScan. Thurston presented evidence that the decision to add ferritin was made at a lower level for legitimate reasons. Thur-ston and Isola testified that at the general managers meeting in early 1988, management decided to allow individual laboratories to determine whether to add ferritin to the LabScan. There was also testimony that Damon’s sales force made requests to the sales managers and general managers of individual labs to add ferritin to the LabScan, on the grounds that the new test would make the panel more competitive. 2. Differential Charging for Ferritin Test (a) Thurston’s Knowledge of Price Differential Medicare provided one reimbursement for the ferritin test and a separate reimbursement for the LabScan panel. In 1988, for example, Medicare paid $24.05 for a LabScan and $20.86 for a ferritin test. Thurston’s labs submitted to him capital expenditure requests (CERs) to help with the decision whether to bundle the tests. CERs were financial projections comparing the additional revenues (from Medicare reimbursements) and costs (partly from the purchase of new equipment) that would result from the addition of ferritin to the LabScan. These CERs assumed there would be no increase in the charge to doctors and patients for a LabScan with ferritin. Thurston discussed and then signed the CERs. They projected losses of revenue from doctors and patients (who would receive free ferritin tests) and massive increases in revenue from Medicare reimbursements. One lab, for example, projected that the number of ferritin tests performed for Medicare beneficiaries would grow from 25 per month to 1,946 per month — increasing revenues by $10,308 per month. Overall, Thurston’s labs projected that they would increase their Medicare reimbursements by approximately $1.16 million per year by adding ferritin to the LabScan; if multiplied by five years, this increase would be $5.8 million. The capital expenditure requests were best-case projections, assuming that every doctor who would otherwise have ordered a LabScan without ferritin would now order a LabScan with ferritin. Damon sought reimbursement (at different rates) for ferritin tests from Medicare, CHAMPUS, and other insurers, but provided them for free to doctors and patients. Thurston said he was unaware doctors were not being charged. But Thurston approved the decision by a number of his labs to adopt this policy of differential pricing depending on the client, and the CERs assumed no price increase for doctors and patients. Thur-ston instructed a subordinate to add ferri-tin at no charge to doctors; Thurston was present at a meeting in which another executive announced the no-charge policy; and Thurston received a memo saying that there was no price increase for doctors at a lab he oversaw. Witnesses testified that physicians are highly price-sensitive about charges for lab work and might object to the automatic inclusion of ferritin unless the test were provided for free. At least initially, many or all of Thurston’s labs increased the charge to private insurance companies for a LabScan based on the addition of ferritin. However, any price increase for private insurers on the expanded LabScan was much smaller than the cost increase to Medicare. Thurston presented evidence that two of his labs did increase the price for physicians; that this price increase provoked complaints by doctors; that he was informed of these complaints; and that he responded that doctors could not obtain a LabScan with ferritin unless they paid the higher price. He also presented evidence that the no-charge approach was a deviation from corporate policy and was initiated by the heads of individual labs; that he did not find out about any price differential until years after it had been implemented; and that he sought to correct any price differential as soon as he discovered it. Thurston testified that he did not authorize, condone, or ratify a decision not to increase the price of the LabScan based on the addition of ferritin. (b) Concealing Pnce Differential from Doctors Thurston’s labs took steps to conceal from doctors that, in addition to the Lab-Scan charge, Medicare would pay an extra fee for the ferritin tests. No letters were sent to doctors advising them that Medicare would be charged separately for the ferritin test. To the contrary, in Apx-il and May 1988, letters were sent to physicians notifying them that the ferritin test was going to be automatically added at “no extra cost.” In Newbury Park, stickers were also printed and added to physicians’ brochures and directories saying, “Ferritin Automatically Included at No Charge.” Of course, as to Medicare patients, these statements were untrue. Thurston testified that he did not pre-approve the letters or stickers saying fer-ritin would be provided free to customers. Thurston portrayed himself as a hands-off manager who trusted subordinates to build in a charge to physicians and to accurately promote the expanded LabScan. Several physicians testified they were initially unaware that Damon charged Medicare for the ferritin component of the expanded LabScan. A number of Damon customers protested, and even switched labs, when they belatedly discovered Damon was charging Medicare for ferritin tests conducted as part of the LabScan. When doctors told Damon sales representatives that they did not need the additional test, some were told that a LabScan without ferritin would cost more than a LabScan with ferritin. A major client referred Damon to a newspaper article criticizing the practice of bundling tests into panels and profiles, and warned Damon that its failure to educate doctors about the composition of and alternatives to its panels would subject it to ongoing criticism. Thurston was informed of these complaints, which were written up in monthly management reports he received in 1989 and 1990. Despite these complaints, Thurston did not cause a letter to be sent to doctors informing them of the extra charge for ferritin to Medicare. Damon also collaborated with HMOs operating on a capitation basis (i.e., paying a flat monthly fee for each HMO member) to reduce utilization of bundled tests such as the LabScan. For example, one of Thur-ston’s labs added to its HMO requisition form a checkbox for a LabScan without ferritin. Damon made no such effort to assist Medicare. 3. Availability of LabScan Without Ferritin Thurston instructed subordinates to take specific steps that hid the fact that the LabScan could be ordered without fer-ritin and made it difficult for doctors to order the LabScan separately. For example, Thurston told the general manager of the Newbury Park lab not to advertise or promote the fact that doctors could still order a LabScan without ferritin. Similarly, Thurston helped make the decision to omit, from the Newbury Park letter announcing to doctors the addition of ferritin to the LabScan, the test code for ordering a LabScan without ferritin. The standard requisition forms used by Thurston’s labs did not change following the addition of ferritin to the LabScan; there was a checkbox for the “LabScan” (which now included ferritin) but no box for the “Lab-Scan without ferritin” or the “LabScan with ferritin.” During the conspiracy period, there were at least two ways for doctors to order the LabScan without ferritin. They could handwrite such an order on a standard requisition form, or they could request and obtain a customized requisition form with a checkbox for the LabScan without ferritin. A number of doctors and institutions took advantage of these options. Thurston testified that he did not tell anyone to conceal or refrain from advertising that the LabScan could be ordered without ferritin. There was testimony that, after Thurston learned in 1991 that the front of a requisition form used by one of his labs did not disclose the inclusion of ferritin in the LabScan, he instructed a subordinate to list it on the front and the form was so amended. 4. Apolipoprotein Damon also added an apolipoprotein test to its coronary risk panel in 1989. Before apolipoprotein was added to the panel, clients ordered it very rarely. The standard requisition forms Damon used after adding apolipoprotein to the panel also did not offer a checkbox for a coronary risk panel without apolipoprotein. Following the addition of apolipoprotein, the cost to doctors of the coronary risk panel increased by five dollars. Medicare reimbursed apolipoprotein separately at a rate well above five dollars. Thurston participated in the decisions about the addition, the requisition forms for, and the extra charges for apolipoprotein. D. Thurston’s “Reasonable Interpretation” as Evidence of Lack of Criminal Intent Thurston presented evidence that from 1988 to 1993 it was an industry-wide practice for labs to rely on the doctors who ordered tests to make determinations of medical necessity. Because, prior to 1994, doctors did not normally share their diagnoses with labs, labs did not have the information required to gauge medical necessity. Damon employees did not believe that they certified tests as medically necessary when they submitted HCFA 1500 forms with Medicare bills. Instead, they believed it was the doctors who ordered LabScans who made the certification. Two expert witnesses testified that it was appropriate in 1988 to include ferritin in a blood chemistry panel. The bundling of different tests into panels was lawful under Medicare regulations. There was testimony that technological changes during the 1980s made it possible to automate the ferritin test, which presumably made it much cheaper to conduct. Witnesses for both sides testified that by 1988 some of Damon’s competitors offered ferritin as part of their blood chemistry panel and so Damon added the test to stay competitive. II. PROCEDURAL HISTORY A. Indictment and Trial On January 22, 1998, a thirty-nine paragraph single-count indictment charged Thurston and three other former Damon executives — Joseph Isola, Beño Kon, and Gerald Cullen — with conspiring to defraud HCFA in violation of 18 U.S.C. § 371 by causing doctors to order unnecessary tests by adding a test for ferritin to a preexisting panel of diagnostic blood tests, and by adding a test for apolipoproteins to a profile used to assess coronary artery disease. The conspiracy period was from July 1987 to August 1993. Isola, President of Damon, pled no contest and, pursuant to his plea agreement, was sentenced to three years’ probation and a one-hundred dollar special assessment. Kon, Corporate Controller, died during the proceedings. Cullen, Senior Vice President for Operations, was tried before the district court in October 2001 and acquitted at the close of the government’s evidence. In addition, Damon pled guilty on October 11, 1996 to conspiracy to defraud by bundling ferritin with the Lab-Scan and apolipoprotein with the cardiac risk panel. The company was sentenced to pay a $35,273,141 fine, and later entered into a civil settlement under which it paid the United States and the state Medicaid programs an additional $83,756,904. Thurston was tried before a jury in November and December 2001. The trial lasted three weeks. At the close of the government’s evidence, the district court granted Thurston’s motion for judgment of acquittal as to the indictment’s apolipopro-tein allegations, ordered these allegations stricken, and explained to the jury that only the ferritin allegations remained. The jury found Thurston guilty. Thur-ston’s subsequent motions for judgment of acquittal and for a new trial were denied. The procedural rulings at trial that Thur-ston attacks are described below. B. Sentencing The district court sentenced Thurston to three months’ imprisonment (with a judicial recommendation that the term be served in a halfway house), followed by twenty-four months of supervised release (of which the first three months were to be served in home detention). The court imposed a one-hundred dollar special assessment and no fine. The Pre-Sentence Report (PSR) had recommended a Total Offense Level of twenty-six and a Criminal History Category of one. The PSR identified the base offense level as six. It recommended a fourteen-level upward enhancement for an intended loss of at least five million dollar’s; a two-level enhancement for more than minimal planning; and a four-level enhancement for a leadership role. It also suggested that an enhancement for obstruction of justice was appropriate, on the grounds that Thurston perjured himself at his trial. The PSR recommended the statutory maximum term of sixty months’ imprisonment and noted that both the statute and the guideline allowed for a fine. It calculated that Thurston had a net worth of $1,526,904. The defense contested the PSR’s recommendations. Thurston requested a three-level decrease because the substantive offense of defrauding the United States was incomplete; a four-level decrease because Thurston was a minimal participant; and a two-level decrease for acceptance of responsibility. Thurston requested a downward departure on the grounds that he had an extraordinary record of charitable work and community service; that the offense constituted aberrant behavior; and that there was the potential for a large disparity with Isola’s sentence. The government argued that Thurston should receive a two-level upward enhancement for obstruction of justice. Otherwise, it accepted the recommendations of the PSR. The parties agreed that any restitution had been made by Damon, the corporate defendant. The district court sentenced Thurston at a hearing on June 26, 2002 to three months’ imprisonment, a period of supervised release, and no fine. The district court granted a fourteen-level enhancement for the size of the intended loss, see U.S.S.G. § 2Fl.l(b)(0) (1992 version), a four-level enhancement for an aggravated role in the offense, see U.S.S.G. § 3Bl.l(a) (1992 version), and a two-level enhancement for more than minimal planning, see U.S.S.G. § 2F1.1(b)(2) (1992 version). The court did not explicitly rule on the government’s request for an obstruction of justice enhancement, see U.S.S.G. § 3C1.1 (1992 version), or Thurston’s request for a three-level decrease for failure to show completion of the substantive offense, see § 2X1.1 (1992 version). Thurston’s adjusted offense level of twenty-six and criminal history category of one yielded a guidelines sentencing range of sixty-three to seventy-eight months’ imprisonment; this range was trumped by the statutory maximum of sixty months for a violation of 18 U.S.C. § 371. During the hearing, the government confirmed that Thurston had been offered (and had rejected) a plea agreement “along the lines” of the one that Isola had accepted. Over the government’s objections and arguments, the district court then departed downward on the basis of Thurston’s record of charitable work and community service and the disparity between Thurston’s and Isola’s sentences. The court then solicited the government’s recommendation about the extent of the departure. The government responded that, if the court chose to depart, then it should depart no further than the sentencing guidelines for a perjury conviction, which would be an offense level of twenty (six less than that of the underlying offense), for a sentencing range of thirty-three to forty-one months’ imprisonment. The court then departed by at least sixteen levels. The government appealed the sentence and Thurston appealed his conviction and sentence. III. THURSTON’S APPEAL FROM HIS CONVICTION Thurston was convicted under 18 U.S.C. § 371, which provides: If two or more persons conspire either to commit any offense against the United States, or to defraud the United States, or any agency thereof in any manner or for any purpose, and one or more of such persons do any act to effect the object of the conspiracy, each shall be fined under this title or imprisoned not more than five years, or both. A. Statute of Limitations Thurston argues that the trial court erred in not granting his post-verdict Rule 29 motion for acquittal, because the government had not proved an overt act during the limitations period, and in not sua sponte instructing the jury on the statute of limitations. The statute of limitations for 18 U.S.C. § 371 crimes is the general five-year statute of limitations contained in 18 U.S.C. § 3282. Here, that five years ran back from January 22, 1998, the date of the indictment, less the six weeks during which Thurston agreed to toll the limitations period. The government, therefore, had to prove an overt act was done on or after December 11, 1992. The indictment properly alleged at least eight overt acts within the limitations period. Thurston did not raise the defense of statute of limitations either before or at trial, did not request an instruction on the defense, and did not object when the judge instructed without addressing the issue. Thurston first raised the issue by Rule 29 motion after the verdict. The government says Thurston raised the issue too late. There is a preliminary question of when such a motion should be raised, a question affecting our standard of review. “The statute of limitations is a defense and must be asserted on the trial by the defendant in criminal cases.... ” Biddinger v. Comm’r of Police, 245 U.S. 128, 135, 38 S.Ct. 41, 62 L.Ed. 193 (1917). Here the indictment adequately pled facts to establish that the crime was within the limitations period. Thurston was not required to raise the defense before trial under Rule 12(b)(3), Fed. R. Crim P. Nor would it have made sense for him to do so, since the defense depended on what the government proved or failed to prove at trial. In a criminal case a defendant need only plead as to the accusation of guilt in the indictment and need not raise the statute of limitations as an affirmative defense before trial. Thurston mistakes these truisms for an argument that he need not raise the limitations defense at all before the jury delivers a verdict of guilt. The government says Thurston has waived the issue and may not raise it at all. Absent an explicit agreement to waive the defense, we treat the issue as a forfeiture and not a waiver, contrary to the government’s argument. This was not an intentional relinquishment or abandonment of a known right, the definition of a waiver. The issue of failure to assert the defense was viewed as forfeiture in United States v. O’Bryant, 998 F.2d 21, 23 n. 1 (1st Cir.1993). The rule we use — that the defense of statute of limitations must be raised at trial and, if not, is forfeited but not waived — is the rule in most circuits. See United States v. Ross, 77 F.3d 1525, 1536 (7th Cir.1996) (“[I]t is widely accepted that a statute of limitations defense is forfeited if not raised at the trial itself.”) (citing cases). Thurston has indeed forfeited the defense that the government did not prove facts that an overt act occurred within the limitations period. The defense should have been raised at trial. Waiting until after the jury has rendered a verdict of guilt to raise a limitations defense for the first time is inconsistent with the characterization of the statute of limitations as an affirmative defense and would unfairly sandbag the government. Because this was ■ a forfeiture and not a waiver, there is still plain error review available under Fed.R.Crim.P. 52. See United States v. Olano, 507 U.S. 725, 731-32, 113 S.Ct. 1770, 123 L.Ed.2d 508 (1993). Our conclusion is straightforward. The government’s evidence established overt acts by the conspirators within the limitations period, so there was no error at all as to the statute of limitations, much less plain error. The government showed that labs overseen by Thurston (and his co-conspirators) submitted tens of thousands of reimbursement claims to Medicare after December 11,1992 for ferritin tests conducted as part of LabScan orders. The government also presented ample evidence that many of these tests were medically unnecessary and were submitted by doctors unaware that Medicare would be charged separately for ferritin. It was not credible that the ferritin test, ordered less than two percent of the time, suddenly became medically necessary thirty to forty percent of the time within the life span of the conspiracy. Thurston also argues he was entitled to a jury instruction on the limitations point. By failing to request a jury instruction and failing to object to the lack of an instruction, he has forfeited the argument. Fed.R.Crim.P. 30; see United States v. Gallant, 306 F.3d 1181, 1187 (1st Cir.2002) (“[A] party unhappy with a trial court’s jury instruction [must] promptly state the precise objection after the instruction has been given.”). As there was no error, the plain error standard was not met. See Fed.R.Crim.P. 30, 52(b). B. Purported Lack of Criminal Intent and the Requested Reasonable Interpretation Instruction Procedurally, the issue of reasonable interpretation comes up in two ways: denial of Thurston’s Rule 29 motions and denial of his request for a jury instruction. Thurston’s three Rule 29 motions — at the end of the government’s case, at the end of the defense case, and after the verdict — all argued that he lacked the needed criminal intent. Our review is of whether a rational fact finder could conclude, beyond a reasonable doubt, that the government proved the elements of the crime, including intent. United States v. Moran, 312 F.3d 480, 487 (1st Cir.2002). Thurston also requested a jury instruction on reasonable interpretation of the law and preserved his objection to the court’s rejection of the instruction. Thurston argued he reasonably interpreted the law as requiring that the treating physician, not the test lab, certify to the HCFA that the test ordered was medically necessary and reasonable, and that he and the company were entitled to rely on that physician certification. Specifically, Thurston contended that in the relevant time period an independent clinical lab did not violate any aspect of Medicare law by: (1) providing physicians with a panel containing a ferritin test, so long as the physician was given reliable and accurate information about the test and could select the panel without the test; or (2) submitting a reimbursement claim, using a HCFA form or otherwise, so long as a physician ordered the test performed. This is one of his primary arguments on appeal. Whether a particular defense doctrine is germane depends on the crime charged and the facts of the case. This is where Thurston’s argument falters. He argues that he could not have had the needed intent because employees of clinical labs, including Thurston, “were unaware that they were actually certifying the medical necessity of each test performed for every patient” and they could reasonably interpret the law to mean that the treating physician, not the laboratory, made the certification. The argument is beside the point. Thurston was not charged with making a false statement to the United States, the falsity of which turned on an ambiguity in what the law required. Nor was he charged with failing to make a statement required by law in a situation of parallel ambiguity. He was not charged with falsely certifying the medical necessity of the tests ordered. He was charged with the crime - of conspiracy to defraud the United States by inducing physicians through deceit and trickery into certifying tests as medically necessary when the fer-ritin tests were not necessary, thus leading Medicare to pay for unnecessary services. Thurston’s knowledge of the Medicare regulations and of the fact that the ordering physicians would certify the medical necessity of the tests was, ironically, part of the proof of the crime, not a defense. Thurston cannot, under 18 U.S.C. § 371, knowingly conspire to mislead and manipulate doctors into certifying medically unnecessary tests which led to improper payment of Medicare funds and then defend on the basis that he committed no fraud because the doctors, not he, were the ones who certified the tests as necessary. Thurston’s reliance on United States v. Prigmore, 243 F.3d 1 (1st Cir.2001), is misplaced. Prigmore is part of a line of cases charging false statements or failure to make required statements, holding that intent should be measured against an objectively reasonable understanding of the legal requirements to be met, and that a statement is not in fact false or fraudulent if it is based on an objectively reasonable interpretation of that legal requirement. See id. at 17-18. This court first applied this principle in United States v. Rowe, 144 F.3d 15, 21-23 (1st Cir.1998), to a statement that was not in fact false under an objectively reasonable interpretation of a disclosure requirement. In Prigmore, the conspiracy charged was to defraud and impair the functioning of the Food and Drug Administration, in connection with its oversight and regulation of medical devices, through failure to file reports which were required under certain conditions. The fraud alleged was the failure to submit a pre-market approval information supplement to the FDA, but whether such a supplement was required depended on the interpretation of certain regulations. The same conditional requirement was true of certain testing reports. The question was whether defendants could objectively and reasonably understand one regulatory phrase, “affecting the safety or effectiveness of the device,” as being circumscribed by another regulatory phrase, “intended ... conditions of use.” See 243 F.3d at 15. No similar question was presented here. Here, the underlying crime was one of manipulating doctors into making false certifications so Damon could receive unwarranted Medicare payments. There is no material question about ambiguity in the underlying legal requirements and no germane question about the meaning of the law. There was also no issue of lack of fair notice of what the law requires, a concern underlying the Prigmore/Rowe line of cases. A reasonable person knows it is wrong to trick others into doing something wrong that one does not do directly oneself, especially in order to obtain personal gain. The Prigmore doctrine has no application given the crimes charged and the facts involved. Because the nature of the crime charged made the reasonable interpretation doctrine irrelevant, the jury instruction issue disappears. C. Failure of District Court to Respond To Requested Instructions Thurston argues that the district court violated Rule 30, Fed.R.Crim.P., which provides: (a) In General. Any party may request in writing that the court instruct the jury on the law as specified in the request. The request must be made at the close of the evidence or at any earlier time that the court reasonably sets. When the request is made, the requesting party must furnish a copy to every other party. (b) Ruling on a Request. The court must inform the parties before closing arguments how it intends to rule on the requested instructions. Thurston is correct: the district court failed to inform the parties of how it intended to rule on each of the requested instructions before closing arguments, as required by the rule. A description of the interactions of court and counsel sets the stage. Each side submitted extensive requests for instructions, and there were disagreements. The court did resolve the most serious disputes over some of the instructions (for instance, on the reasonable interpretation/ Prigmore question) and told counsel these rulings before closing argument. The court did not, though, review all of the requests. Thurston’s counsel did not object to this silence before giving his closing. Defense counsel did raise an issue after closing, and before the jury was instructed, that he wanted to put on record his specific objections to the government’s requests. He did not say he had been prejudiced in any way by the court’s failure to rule on the requested instructions before he gave his closing. The court replied that it would neither rule on nor hear argument on the proposed instructions. Rather, the court stated its understanding that the appropriate time to object was at the end of the instructions. Nonetheless, it did hear argument on the government’s Request No. 8 (the compelled witness rule), and declined to give the instruction. It also heard argument on the government’s proposed instructions No. 18 (conspiracy); No. 19 (unlawful objectives); and No. 22 (overt act). The only proposed defense instruction called to the court’s attention was No. 24, on missing witnesses. The district court did not, as Rule 30 requires, tell counsel before closing argument its disposition of all of the requested instructions. But counsel for Thurston had an obligation to bring this to the court’s attention before the closing and did not do so. Without addressing the issue of whether Thurston has thus forfeited the Rule 30 argument, we choose to simply evaluate whether defense counsel’s closing argument was adversely affected. See United States v. Owens, 167 F.3d 739, 753 (1st Cir.1999). It was not. One telling indicium that there was no prejudice is that trial counsel did not ever say to the court he would be prejudiced if he had to proceed with his closing without knowing the court’s disposition of the remaining requested instructions. An even more telling indicium of lack of prejudice is that Thurston’s appellate counsel has been unable to identify any specific areas of prejudice occasioned by the trial court’s lapse. While in theory such a lapse could cause prejudice, the most that is argued here is that there was “no detailed reference to important legal concepts regarding criminal conspiracy and the state of mind by which Thurston would be judged.” Appellate counsel does not identify those “important legal concepts,” and we see none. As to Thurston’s state of mind, the trial judge did instruct on the government’s burden to show Thur-ston had a specific intent to participate in the conspiracy and to defraud the United States. The court instructed the jury to consider Thurston individually to determine if he willfully joined the conspiracy. The court, in turn, defined “willfully.” The court also explicitly rejected Thur-ston’s reasonable interpretation instruction, and it instructed on good faith. These circumstances belie any claim of prejudice and Thurston’s claim fails. D. Motion for New Trial Based on Dismissal of Apolipoprotein Charge Thurston and his co-defendants were originally charged with conspiracy to commit fraud as to both the ferritin and the apolipoprotein tests. At the close of the government’s case the court granted Thur-ston’s motion for judgment of acquittal on the apolipoprotein test, struck those references from the indictment, and instructed the jury that this issue was no longer before it. The government did not thereafter refer to this issue. Thurston now argues that the court should have granted Thurston a new trial after the jury returned because the apoli-poprotein evidence irretrievably tainted the trial. The government rejoins that counsel should have raised the issue sooner. Again, we bypass the issue of forfeiture and reject the argument that dismissal of the apolipoprotein charges tainted the proceedings. Thurston’s argument that none of this evidence would have been admitted if the ferritin charges were tried alone is based on an unlikely premise. Where the evidence admitted as to a dismissed count would have been admissible as to a remaining count, the defendant has not suffered prejudice. United States v. Rooney, 37 F.3d 847, 855-56 (2d Cir.1994) (collecting eases); see United States v. Weiner, 3 F.3d 17, 22 (1st Cir.1993) (jury properly considered evidence relating to counts dismissed prior to verdict, since evidence was relevant to remaining counts). The government would have introduced such evidence in any event as relevant to rebut central defense themes that, because Damon had such a decentralized decision-making structure, Thurston was not involved in key decisions. The apolipopro-tein evidence contradicted Thurston’s claims about the extent and consequences of Damon’s decentralized approach to the make-up, pricing, and marketing of its panels. There was little risk of prejudice for other reasons. The government did not mention the apolipoprotein evidence in its closing, and exhibits pertaining only to apolipoprotein were removed before the documents were submitted to the jury. Further, most of the testimony and documentary evidence in the first half of the case, before the court ruled on the Rule 29 motion, dealt with ferritin. These factors further minimized the likelihood of any taint. Thurston’s conviction is affirmed. IV. SENTENCING APPEALS A. Thurston’s Appeal: Loss Calculation Much of Thurston’s guidelines sentence range (sixty-three to seventy-eight months) was driven by the loss calculation. Both the PSR and the government recommended an intended loss figure of more than five million dollars but less than ten million dollars. This resulted in a fourteen-level increase in the base offense level. Thurston, not surprisingly, targets this loss calculation. He makes two arguments. The first is that the government was precluded by a comment to U.S.S.G. § 2F1.1 from ever relying on intended loss unless the government first established what the actual loss was and then established that the intended loss was greater. This is a pure issue of guidelines interpretation, which we review de novo. See United States v. Gonzalez-Alvarez, 277 F.3d 73, 77 (1st Cir.2002). The second argument is that the court’s conclusion had insufficient factual support for a number of reasons, a contention reviewed for clear error. The guidelines interpretation argument turns on a comment that provides: Consistent with the provisions of § 2X1.1 (Attempt, Solicitation or Conspiracy), if an intended loss ■ that the defendant was attempting to inflict can be determined, this figure will be used if it is greater than the actual loss. U.S.S.G. § 2F1.1, cmt. n. 7 (Fraud and Deceit) (1992 version). Thurston argues that because the government did not show actual loss, it cannot turn to intended loss. The argument is simply wrong as a matter of the wording of the comment. The comment directs the use of an intended loss figure when it is greater than the actual loss figure; the comment does not restrict the sentencing court’s ability to rely on intended loss when there is no actual loss calculation available. Defendant’s reading also makes little sense: it may be easier as a matter of proof to show intended loss than actual loss. Conspirators are held accountable for the loss they intend to commit. Finally, it is obvious on these facts that the intended loss was greater than the actual loss — some doctors quit using Damon when, to their disgust, they realized what the scheme was. Thurston next mounts a series of fact-based attacks on the intended loss figure of more than five million dollars. That figure is supported by the capital expenditure requests each of the labs prepared to obtain funding to buy the equipment needed to perform the increased ferritin testing the labs anticipated. Each CER included a financial analysis, one component of which was the estimated new revenue from bundling the tests. Thus, the intended loss calculation was based on the conspirators’ own financial calculations. Thurston argues the CERs, being mere financial projections, are not an adequate basis for an intended loss figure; that the CERs were based on an assumed best-case scenario in which no physician who ordered LabScans would decline to get ferritin tests; that, in any event, not all physicians who ordered the bundled test were tricked into doing so; that the conspiracy was not proven to last five years; that the conspiracy ended before five years had elapsed for several labs, given the different dates on which the labs bundled the tests; and that later, only one lab reported to Thurston. A number of these ai-e quickly dispatched. The jury verdict of guilt disposes of the question of the length of the conspiracy. Thurston’s promotion out of management of three of the labs is irrelevant since he earlier conspired to produce losses intended to go on for years. Thurston also argues that the intended loss had to be reduced under U.S.S.G. § 2X1.1(b)(2) because the government’s proof did not establish that the conspirators had “completed all the acts the conspirators believed necessary on their part for the successful completion of the substantive offense.” Id. There is no merit to the argument. There was successful completion of the offense: the tests were bundled and doctors were misled into ordering unnecessary ferritin tests. The complaints from customers about Damon’s practices were confirmation the scheme had worked. The closer question is the degree of precision the government must reach in showing intended loss. It is true that the CERs set forth best-case scenarios which assumed all doctors would order the bundled test without culling out the ferritin test. If the CERs stood alone, defendant would have a better argument. But they are supplemented by the fact that Damon made it extremely difficult for doctors to cull out the ferritin and order the LabScan without a ferritin test. Further, the conspirators tried to hide from doctors the fact that there was a significant cost to Medicare associated with the bundling. The conspirators were maximizing the probability that all doctors would accept the bundling, without culling and without protest. The fact that the conspirators were not entirely successful in fooling all doctors does not lessen their intent. We have noted before that intended loss does not have to be determined with precision; the court needs only to make a reasonable estimate in light of the available information. United States v. Blastos, 258 F.3d 25, 30 (1st Cir.2001). There was good evidence of intent and some “prospect of success” for the fraud to reap over five million dollars, and that is all that the case law requires. United States v. Orlando-Figueroa, 229 F.3d 33, 48 (1st Cir.2000). The best-case CER projection was a loss to Medicare over the charged five-year life of the conspiracy of $5,800,230. It was reasonable for the district court to estimate that the intended loss exceeded five million dollars, even allowing for the one to two percent normal order rate for ferritin tests. The government met its burden and Thurston offered little in rebuttal except his protestations of innocence. There was no clear error. B. Government’s Appeal Given an intended loss of five million dollars, Thurston’s crime led to the statutory maximum sentence of five years. The district court departed downward from the Guidelines range by sixteen levels, however, sentencing Thurston to three months’ imprisonment to be followed by twenty-four months of supervised release. The court imposed no fine and recommended that the term of imprisonment be served in a halfway house. 1. Standard of Review: The PROTECT Act On April 30, 2003, section 401 of the PROTECT Act became effective. That section changes the applicable standard of review for certain issues in appeals from departures from the sentencing guidelines. Section 401 amends 18 U.S.C. § 3742(e), which now provides: (e) Consideration. — Upon review of the record, the court of appeals shall determine whether the sentence— (3) is outside the applicable guideline range, and (A) the district court failed to provide the written statement of reasons required by section 3553(c); (B) the sentence departs from the applicable guideline range based on a factor that— (i) does not advance the objectives set forth in section 3553(a)(2); or (ii) is not authorized under section 3553(b); or (iii) is not justified by the facts of the case; or (C) the sentence departs to an unreasonable degree from the applicable guidelines range, having regard for the factors to be considered in imposing a sentence, as set forth in section 3553(a) of this title and the reasons for the imposition of the particular sentence, as stated by the district court pursuant to the provisions of section 3553(c);.... The court of appeals shall give due regard to the opportunity of the district court to judge the credibility of the witnesses, and shall accept the findings of fact of the district court unless they are clearly erroneous and, except with respect to determinations under subsection (3) (A) or (S)(B), shall give due deference to the district court’s application of the guidelines to the facts. With respect to determinations under subsection (3) (A) or (3)(B), the court of appeals shall review de novo the district court’s application of the guidelines to the facts. (emphasis added). This changed the law. Under Koon v. United States, 518 U.S. 81, 116 S.Ct. 2035, 135 L.Ed.2d 392 (1996), the courts of appeals were not to review a departure decision de novo, but were to ask whether the sentencing court abused its discretion in granting the departure. Id. at 91, 96-100, 116 S.Ct. 2035. In appeals from sentencing departures, we had, before the PROTECT Act, engaged in a three-part review: “(1) we determine[d] whether the stated ground for departure [was] theoretically permissible under the guidelines; (2) if so, we examine[d] the record to assess whether there [was] adequate factual support; and (3) we determine^] the appropriateness of the degree of departure.” United States v. Bogdan, 302 F.3d 12, 16 (1st Cir.2002). Whether the stated ground for departure was theoretically permissible— the first part — was a question of law reviewed de novo. United States v. Bradstreet, 207 F.3d 76, 81 (1st Cir.2000); see also United States v. Diaz, 285 F.3d 92, 97-98 (1st Cir.2002). Under Koon, our review under the remaining two parts was for abuse of discretion only. See Koon, 518 U.S. at 96-100, 116 S.Ct. 2035; United States v. Lujan, 324 F.3d 27, 31 n. 5 (1st Cir.2003); United States v. Martin, 221 F.3d 52, 55 (1st Cir.2000). After the PROTECT Act, the statute requires de novo review not merely of the ultimate decision to depart, but also of “the district court’s application of the guidelines to the facts.” § 3742(e). If this court agrees that the decision to depart was justified under the guidelines, however, the extent of the departure granted by the district court is reviewed deferentially, just as it was prior to the PROTECT Act. Id.; United States v. Frazier, 340 F.3d 5, 14 n. 4 (1st Cir.2003); see also United States v. Mallon, 345 F.3d 943, 946 (7th Cir.2003); United States v. Jones, 332 F.3d 1294, 1300(10th Cir.2003). (a) Retroactivity Thurston argues that the PROTECT Act should not be interpreted to apply to this case and that, if it does apply, it is retroactive and invalid. He makes two statutory intent arguments: (1) that the internal structure of the statute means it should not be applied to cases already pending on appeal; and (2) that the presumption against retroactivity should apply- First, Thurston argues that Congress meant application of the de novo review provisions in the PROTECT Act to be deferred until appeals arise from sentences entered after the Act became effective. This is evident, Thurston says, since the Act imposed a new requirement for the district judge to give a written statement of reasons. From this, Thurston argues, all provisions of the Act were meant to apply only to post-Act sentencing. The argument is plausible, but we are unpersuaded. Even before the PROTECT Act, a trial court was required to give some reasons, though not necessarily in writing, for a downward departure. See 18 U.S.C. § 3553(c)(pre-PROTECT Act version); United States v. Sclamo, 997 F.2d 970, 973 (1st Cir.1993) (discussing discouraged ground for departure); United States v. DeMasi, 40 F.3d 1306, 1324 (1st Cir.1994) (same). A requirement that this statement of reasons be written, rather than oral, has no particular connection to the appellate standard of review. Although the Act does not expressly say that its de novo review provision applies to pending appeals, it does give an effective date of April 30, 2003. The effective date of a statute does not by itself establish that it has any application to conduct that occurred at an earlier date. See INS v. St. Cyr, 533 U.S. 289, 317, 121 S.Ct. 2271, 150 L.Ed.2d 347 (2001) (quoting Landgraf v. USI Film Prods., 511 U.S. 244, 257, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994)). Still, we agree with the Eighth Circuit that the new statute applies to appeals pending as of the effective date of the statute. See United States v. Aguilar-Lopez, 329 F.3d 960, 962-63 (8th Cir.2003). Subject to constitutionally based retroactivity concerns, it is certainly within Congress’s power to change a standard of review. See, e.g., Hines v. Sec’y of Dep’t of Health & Human Servs., 940 F.2d 1518, 1523 (Fed.Cir.1991); Consumers Union of U.S. v. FTC, 801 F.2d 417 (D.C.Cir.1986); cf. Bierce v. Waterhouse, 219 U.S. 320, 336-37, 31 S.Ct. 241, 55 L.Ed. 237 (1911). Much of the conduct regulated by this part of the PROTECT Act is that of the courts of appeals (and indirectly, the district courts now under closer scrutiny), and that involves conduct dating from April 30, 2003 forward. Thurston’s fall-back argument is that applying a changed standard of review to a case already on appeal would have an impermissible effect on him under the Supreme Court’s retroactivity jurisprudence. See Landgraf, 511 U.S. at 264, 114 S.Ct. 1483. Not so. The change of a standard of appellate review is one in procedure for the courts; procedural changes that do not affect substantial rights are not usually considered impermissibly retroactive. This legislation is little different than the Supreme Court’s changing the standard of review by directing the courts of appeals to decide ultimate Fourth Amendment questions de novo. Cf. Ornelas v. United States, 517 U.S. 690, 697, 116 S.Ct. 1657, 134 L.Ed.2d 911 (1996). The PROTECT Act’s alteration of the appellate standard of review upsets no legitimate reliance interest by a defendant; it could not have induced alteration of the behavior that led to the crime. We see no unfairness to defendants in Congress’s requiring a closer look by appellate courts at whether a district court committed an error in deciding that the guidelines permitted a departure. It is the substance of the sentencing rules, both in the Guidelines and in the underlying statutes, that affects defendants. (b) Separation of Powers Thurston makes a cursory argument that the PROTECT Act presents serious constitutional separation-of-powers questions. At the request of the Senate, the Chief Justice, expressing the views of the U.S. Judicial Conference, did advise the Senate of the Conference’s opposition to portions of the bill, including alteration of the standard of review. See Letter from Chief Justice William H. Rehnquist to Senator Patrick Leahy (undated), available at http://www.nacdl.org/public.nsP 2cdd02b415ea3a64852566d6000daa79/ departures/$FILE/Rehnquist_Ietter.pdf. The U.S. Sentencing Commission requested that Congress not act until the Commission had the opportunity to analyze data and study the matter. See Letter from Judge Diana Murphy, Chair of the U.S. Sentencing Commission, et al., to Senators Orrin Hatch and Patrick Leahy (April 2, 2003), available at http://www.nacdl.org/public.nsf/ 2cdd02b415ea3a64852566d6000daa79/ departures/$FILE/$TJstcg_comm_ current.pdf. But judicial opposition to legislation on policy grounds is one thing; unconstitutionality is quite another. No real theory of unconstitutionality has been presented by this appeal, and so the issue is waived. See United States v. Zannino, 895 F.2d 1, 17 (1st Cir.1990). (c) Scope ofDe Novo Review A different argument was advanced in the petition for rehearing concerning the scope of de novo review under the PROTECT Act. The final sentence of § 3742(e) requires the court of appeals to review de novo all determinations by the district court “under subsection (3)(A) or (3)(B),” the texts of which are set forth above. In his petition for rehearing, Thur-ston, ably assisted by amicus, argues that subparagraphs (3)(B)(i) and (3)(B)(ii) apply only in cases involving “unmentioned” departures, ie., departures on grounds not expressly considered by the Sentencing Commission. On this theory, because good works departures are specifically authorized (albeit discouraged) by the guidelines, this court is not authorized to ask whether, in Thurston’s case, such a departure (i) does not advanc