OPINION

KAPLAN, District Judge.

Great research libraries are repositories of our social, cultural, and scientific heritage. Their rare books and manuscripts are vital to understanding the world and often are irreplaceable objects of study for scholars who add to our knowledge of ourselves and our environment.

The defendant in this case stole hundreds of rare manuscripts and documents from the library of Columbia University. The matter now is before the Court in consequence of the Court's indication that it would consider an upward departure under the Sentencing Guidelines in view of the nature of the materials stolen.

I.

Daniel Spiegelman pleaded guilty on April 17, 1997 to a three count information that charged him with (1) interstate and foreign transportation of stolen goods, (2) making a false statement in a passport application for the purpose of obtaining a passport for his own use, and (3) shipping and transporting firearms in interstate commerce after having been convicted of a crime punishable by more than one year in prison. The plea agreement stipulated, inter alia, that the loss caused by the defendant's theft was approximately $1.3 million, that the applicable Guidelines offense level is 18, and that the criminal history category is II. Based on those assumptions, the sentencing range would be 30 to 37 months imprisonment absent any departure.

The initial presentence report, dated June 11, 1997, adopted the stipulated $1.3 million loss amount set forth in the plea agreement and recommended that the Court adopt the offense level and criminal history category stipulations contained in the plea agreement. The presentence report, however, was accompanied by a letter from Jean Ashton, the director of the Columbia University Rare Book and Manuscript Library, which argued that the defendant's crime had a "devastating effect" on Columbia and the worldwide academic community far greater than the financial loss because of its impact on scholarship.

Spiegelman first appeared for sentencing on June 18, 1997 at which time the Court adopted the factual findings and Guideline computation in the presentence report. It indicated also that it was considering an upward departure under U.S.S.G. § 5K2.5 on the ground that "the loss table does not adequately reflect the gravity of the crime." After hearing argument, it adjourned the sentencing to afford defendant an opportunity to address the proposed departure issue in more detail.

Spiegelman appeared again for sentencing on July 30, 1997, at which time the departure issue was argued further. The Court rejected defendant's legal argument that an upward departure would be inappropriate as a matter of law. Defendant, however, indicated for the first time a desire to submit evidence on the issue whether the stolen items had unique or special scholarly value, so the Court again adjourned the sentencing in order to give defendant time to develop his factual argument. In an effort to determine whether to hold a Fatico hearing, the Court directed defendant to provide a statement "listing by name each witness it proposes to call at a Fatico hearing and a summary of the witness's proposed testimony and the basis for the witness's giving that testimony." The Court directed also that the government submit a similar statement "including any evidence that Columbia University wishes to present."

On August 14, 1997, defendant requested a hearing to "present and resolve factual issues as to [the] value to scholarship of [the] stolen documents" and to cross-examine Columbia's Ms. Ashton. The government, in compliance with the Court's direction, then reported that Ms. Ashton and eight other scholars were prepared to testify at any Fatico hearing.

Following the receipt of these communications, the Court directed the Probation Department to obtain written statements of the proposed testimony of these eight individuals. On November 20, 1997, the Probation Department issued a supplemental presentence report which attached and summarized letters from each of the eight scholars. Defendant responded by reiterating his desire to cross-examine Ms. Ashton.

Upon receipt of all of these materials, the Court granted Spiegelman's request for a Fatico hearing for the purpose of affording him the opportunity to (1) cross-examine Ms. Ashton and Professor George Saliba, author of one of the eight letters submitted with the revised presentence report, and (2) present two witnesses of his own. The hearing was conducted on March 20, 1998. As Professor Saliba was not available for the hearing, the Court granted Spiegelman's motion to disregard his letter. Spiegelman proved unable to procure the appearance of his proposed witnesses. Accordingly, only Ms. Ashton testified.

At the conclusion of the Fatico hearing, the Court granted Spiegelman's request for ten days in which to submit another brief. Spiegelman filed the brief 31 days later. The brief argued for the first time that Section 5K2.5 does not permit a departure in these circumstances and, moreover, that departure under Section 5K2.0 would be inappropriate. Although Spiegelman's brief thus reflected his awareness that the Court would consider departing under Section 5K2.0 as well as 5K2.5, the Court gave formal notice to that effect on April 20, 1998.

II.

The Stolen Items

Spiegelman stole, among other things:

• seventeen medieval and Renaissance manuscripts dating from 1160 to 1550, including Euclid's Elementa (1300), three Books of Hours with illuminations (1425-1490), a late fourteenth century manuscript of Roman de la Rose by Guillaume de Lorris and Jean de Meung, and two papal bulls (1160 and 1202);

• eight Arabic and Persian manuscripts dating from around 900 to 1887, including several Korans and a volume of secular Persian poetry;

• three incunabula: Compilatio de Astrorum Scientia by Leopoldus, Dux Austriae (1489), a 1493 edition of Liber Chronicarum, also known as the Nuremberg Chronicle, and Seneca's Proverbia (1493);

• twenty-six medieval, Renaissance, and early modern documents dating from 1122 to 1789, including property deeds, indentures, bills of sale, tax documents, and wills;

• 284 historical maps dating from 1628 to 1891, including Danckerts, Novi Belgii no Aeque Angliae . . . Tabula (1700), and 237 individual maps razored out of Blaeu's Atlas Major (circa 1667, German text edition);

• twenty-six Presidential letters and documents dating from 1791 to 1918, including six George Washington letters (four to John Jay), eight letters from John Adams to William Tudor, and twelve letters from other presidents including John Quincy Adams, Andrew Jackson, Abraham Lincoln, and Woodrow Wilson; and

• 133 letters and documents from and relating to Thomas Edison dating from 1860 to 1903, including contracts, patent assignments, and correspondence concerning the development of the telegraph and the expansion of the telegraph network.

Many of these items have not been recovered. Spiegelman led the FBI to others following his arrest.

The Impact of the Thefts

The theft of these items concededly caused economic loss to their owner, Columbia University. But the theft had an impact different in kind from a loss of money or other easily replaceable property, for these materials have value to the Columbia academic community and other scholars and, through them, to society at large that cannot be measured in economic terms alone.

Ms. Ashton broadly and aptly summarized the importance of materials of this nature in a statement which the Court fully credits:

"The very existence of rare books and manuscripts provides the basis for new discovery and interpretation in almost every area of study, and can continue to do so only for as long as these treasures remain preserved and protected. As Roger Stoddard, curator of rare books in the Harvard College Library, aptly stated . . . `we need all the evidence that we can get [to understand history] — scarce as it is, and we depend on the continuing accessibility of old books and manuscripts, so we can test the accuracy of new interpretations. . . . All depends completely on the maintenance and security of library collections: destroy, mutilate, steal, or hide the books and manuscripts and you frustrate the development of knowledge and the free interchange of scholarship and teaching.'"

As Professor Blackmar pointed out, "[t]he value of rare books and manuscripts derives from their availability to scholars whose research entails analyzing both individual, uniquely significant documents and historical patterns that emerge from multiple documents." The negative effect that theft can have on the atmosphere and security policies of rare book collections, moreover, is substantial because it brings security concerns to the fore at the expense of accessibility.

The foregoing merely touches the surface of the extensive evidence before the Court that the theft of rare books and manuscripts deprives Columbia and similar institutions, their faculty and students, and other scholars of raw materials important to the advancement of human knowledge and thus deprives all of us of the benefits of their insights and discoveries. Nonetheless, Spiegelman argues that his thefts had no such impact for a number of reasons.

1. Evidence of Specific Injury

One of Spiegelman's refrains, variously expressed, is the suggestion that there is no proof that there was any specifically identifiable harm to scholarship as a result of his theft of these particular items, especially as some of them were recovered following Spiegelman's arrest. In the very narrowest sense he is right. It is impossible to say with an acceptable degree of confidence that the loss of a particular book or manuscript resulted in an identifiable loss of a specific intellectual insight or historical contribution, much less that the interim loss of those materials ultimately recovered did so during the period in which they were unavailable. It does not lie within human competence to say what would have occurred had the thefts not taken place — any more than we can say what discoveries would have been made or insights achieved but for the loss of knowledge occasioned by the destruction of the Library Alexandria, the rampant looting of European art treasures during World War II, or the infamous book-burnings conducted by the Nazi regime in Germany.

In the last analysis, then, by taking raw materials of importance to scholars and the academic community, Spiegelman knowingly risked causing substantial harm of a kind which is inherently unknowable, unquantifiable, and substantially non-monetary in character. The fact that we cannot know with certainty what new knowledge would have come from which artifact absent Spiegelman's actions, far from supporting Spiegelman's position, supports the view that a monetary measure alone cannot possibly account fully for the seriousness of his offense.

2. Attacks on the Significance of Particular Items

Spiegelman next contends that certain of the items stolen are of little importance. He begins with the assertions that the presidential papers are of "limited scholarly value" because they deal with "appointments to minor positions" and that "[l]ittle use has been made of them, even though they have existed for some two hundred years." He attacks the significance of the Edison papers in much the same vein. The problem with these broad assertions, however, is that the competent and persuasive evidence all is to the contrary.

Professor Blackmar, for example, explained that the Edison patent transfers are salient to the history of intellectual property and technology, "both fields of scholarship that are `taking off,'" and are "exactly the kind of collection that I have asked students to examine and make sense of in my undergraduate seminar on the American Gilded Age (which I teach in the Rare Books and Manuscript Library)." She went on to explain that presidential letters concerning even minor appointments have definite scholarly value:

"[G]iven that President James Garfield was assassinated over the question of patronage (and by someone disappointed at not having received an appointment to a `minor position'), the substantive questions that attach to such letters are far from trivial. Indeed, understanding the distribution of patronage is vital to understanding nineteenth century American politics. * * * Whether it be historians investigating the service of a particular president or a scholar seeking to analyze the conventions of presidential appointments at the lower levels, there is little reason to assume that these presidential letters have ceased to be of value for scholarly research, assuming that they are available for that purpose in the Rare Books and Manuscript Library."

The Court therefore finds that the Edison and presidential papers have significant scholarly value.

Spiegelman asserts also that the various medieval manuscripts are unimportant and do not differ meaningfully in physical characteristics from other available materials. This broad assertion, however, is contradicted by Professor Somerville, who discussed the significance of both the intellectual content and physical characteristics of the stolen papal bulls, Plimpton MS 278 (of Pope Innocent III, dated 1202) and Smith Western MS 8 (of Pope Nicholas III, dated 1279). According to Somerville, Plimpton MS 278

"represents a rare witness from the chancery for a period when the official chancery register of correspondence for this pivotal reign (Innocent III [1198-1216] is sometimes counted as the most important medieval pope) is missing. Furthermore, so few medieval originals survive, that each one can provide information about the chancery officials who prepared these documents, and about the form and style which the letters assumed, e.g., how the bulls were attached, what sort of script was used, what signs of authentication appear thereupon."

Somerville highlights also the significance of having the bulls in the university collection, where they were available to students and scholars alike and could be placed "in `conversation'" with related materials.

The foregoing points are merely representative of Spiegelman's attacks on the significance of the materials in question. They suffice, however, to demonstrate their vacuity. In each instance, unsubstantiated assertions by Spiegelman confront detailed refutation by qualified and, indeed, renowned experts in the relevant fields. The Court credits the experts and rejects Spiegelman's position. The stolen materials have substantial scholarly value.

3. The "Other Copies" Arguments.

Spiegelman has endeavored to minimize the harm that he has done by contending that there has been no loss to scholarship, even with respect to materials that have not been recovered, because the materials either had been photocopied before they were stolen or, while perhaps rare, were not unique. The argument has substantial difficulties.

To begin with, many of the materials that Spiegelman stole are unique. No copies exist.

Second, it is important to recognize that most of the rare books and manuscripts at issue here have two quite distinct types of value to scholars, depending upon their particular interests. Substantially all of the materials are important for their intellectual content. Presidential letters concerning appointments, for example, may provide insight into the patronage process. To the extent that good photocopies exist, the copies may permit study of the content to the same extent as the originals. As several witnesses described, however, other materials are of value for their physical characteristics as well as their content. The utility of a copy in such a case depends upon the extent to which the copy reproduces the relevant characteristics of the original. The evidence at the Fatico hearing convincingly established that photocopies of a substantial part of the stolen material, even if they existed, would be inadequate because of their inability to reproduce physical and other characteristics of the originals which are important to those who study them.

Nor is the impact of the loss even of those materials which are rare rather than unique eliminated by the existence of other more or less contemporaneous copies. The mass production of identical books to which we are accustomed in the late twentieth century is a comparatively recent development. As Ms. Ashton demonstrated at the hearing, even two copies of the Nuremberg Chronicles, both printed in 1483 by the same printer, have important differences, and the differences are significant to those who study the development of the book.

In summary, Spiegelman's arguments regarding copies do not undercut in any meaningful way the Court's finding that the books and documents Spiegelman stole are of substantial importance to scholarship. Copies exist in few if any cases. Even if they did, they frequently would be inadequate substitutes for the originals. Indeed, faced with the overwhelming evidence, Spiegelman stipulated at the Fatico hearing that photocopies, even if they existed, would not be adequate substitutes for the original medieval and Renaissance manuscripts. The Inadequacy of Economic Value as a Measure of the Loss

As will be discussed in greater detail below, the extent to which the economic value of the material stolen by Spiegelman adequately captures the gravity of his offense is central to the issue whether a departure is appropriate here. It therefore is necessary to consider whether the $1.3 million economic value stipulated by the parties adequately reflects the loss to scholarship and to society caused or risked by Spiegelman's actions.

To begin with, the evidence before the Court persuasively demonstrates that any relationship between the economic loss — in substance, appraisers' views of what the stolen materials would bring if offered for sale — and the importance and value of those materials to scholarship would be purely accidental. Ms. Ashton explained the distinction between the economic value reflected in an item's price and the scholarly value of that item in the following manner:

"The values assigned by appraisers or dealers reflect the fluctuations of demand in an active marketplace of collectors and investors who often speculate in intellectual properties as they do in the financial market; these values do not, and cannot encompass the real or potential worth of any single unique document for the scholar who uses it to interpret the past. . . . A single letter might be had for several hundred dollars at auction. In a library, viewed by a reader who brought to his reading decades of informed scholarship, that single piece of paper becomes priceless, its value enhanced by its presence in an institution where it is accessible for interpretation and use. The importance of such a letter may be magnified by the presence of hundreds of other similar documents that allow the focused researcher to discern peculiarities or discover patterns."

Ms. Ashton elaborated on the point at the hearing:

"The auction or appraised value is a value that is put on it by people who deal in the buying and selling of manuscripts, and that valuation fluctuates according to what happens to be fashionable at the time.

"It may be parallel to what scholars would value it, but scholarly value is entirely separate since a document that is worth $10 on the market may well be the key to an argument that is made by a scholar which influences thousands of people. Conversely, a book that is elaborately illustrated and appears to be extremely valuable and appeals to collectors may have relatively little scholarly value at the time.

"Also, scholarly value is calculated in terms of long-term life in a library. In 50 or 60 or a hundred years, things are rediscovered. They gain and lose value in an entirely different way. The market is slow to respond.

"An example might be women's diaries, papers dealing with slave transactions which were [of] very little value 50, 60, a hundred years ago, and now come [to] command very high prices in the market, largely because they were rediscovered by scholars at a time when they had low scholarly value, and they were collected by libraries who believed in keeping the artifacts of the past.

* * * * * *

"Materials are reused and rediscovered every 20 or 30 years. To give an analogy, a diary written by a midwife in Maine in the 18th Century was looked at many times by scholars and ignored as something trivial. A contemporary historian reexamined it and found it a great window into understanding colonial life in the 18th Century and published it, and it has become a best seller in the historical community."

In other words, what appeals to collectors and those who drive demand in the market for rare books and manuscripts often is quite different from the characteristics of material that may be of interest to scholars and yield important information advancing knowledge.

Ms. Ashton's testimony was echoed by other experts. Professor Dutschke explained that booksellers often cannot anticipate scholarly value that may be inherent in a particular item and therefore cannot reflect that value in price. And Roger Stoddard of Harvard made yet another point when he wrote:

"The figure of 1.6 million dollars has been mentioned for the Columbia University Library thefts, but we are not speaking of stocks from the Columbia investment portfolio — a simple 1.6 million — but stocks from its teaching and research portfolio. A book or document appraised for a few hundred dollars could make the difference in an academic career by providing material for a dissertation, article, or book. One tiny piece of evidence well-studied can bring new insight to a scholar. Such a piece can be appraised only in terms of that career."

Nor is the failure of the appraisals to capture the intangible value of these materials the only difficulty with giving conclusive weight to this economic measure in determining the severity of Spiegelman's actions. While the other matters discussed in this opinion are more than sufficient to justify the departure decision, it bears mention that the valuation in this case, although the Court has adopted it as a starting point for want of any alternative, is quite unsatisfactory. Ms. Ashton, who evidently coordinated the task of obtaining appraisals, testified that the appraisers who were contacted mostly were unwilling "to give detailed appraisals because the material was not there for them to examine." Spiegelman's counsel, in a letter to the Probation Department, wrote that values placed on some of these items were purely theoretical because it is impossible to sell them. It therefore is not surprising that counsel for both sides made it plain that the $1.3 million figure is essentially a ballpark guess necessitated by the fact that reliable valuations were not available.

Spiegelman's principal response to all of this is a contention that the Association of College and Research Libraries ("ACRL") measures loss caused by library thefts exclusively in monetary terms, from which Spiegelman would draw the conclusion that library thefts cause no harm to scholarship or other non-monetary injury or, in any case, that any such impact is unworthy of consideration.

The basis for Spiegelman's argument is a model statute — adopted nowhere, so far as the record discloses — that is attached to a set of ACRL guidelines for dealing with library theft. The model statute proposes that theft, mutilation or concealment of library property be made a crime, the grade of which would be determined by reference to the fair market value of the property. This makes perfect sense for most library thefts, which involve the loss of mass produced and readily replaced materials. In any case, however, the fact that the ACRL saw fit to craft model legislation that draws the grossest possible distinctions in the severity of such crimes on the basis of fair market value, however, is a far cry from the conclusions that Spiegelman asks the Court to reach. The Court declines to find that the ACRL measures the gravity of the loss or destruction of materials such as those at issue here purely in terms of their market value. Even if it did, the Court would be unimpressed with its logic.

For all of these reasons, the Court finds that the $1.3 million economic value placed upon the materials stolen by Spiegelman does not adequately measure the extent of the loss. While the precise details and extent of the non-monetary loss in this case are unknowable, Spiegelman knowingly risked non-monetary loss of the nature described above. Moreover, bearing in mind that the standard of proof here is a preponderance of the evidence, the Court finds that the theft of these materials caused non-monetary loss to Columbia, scholarship, and to society as a result of the inability of scholars to use the stolen materials.

III.

Having thus determined the facts, the Court ordinarily would proceed to consider the proposed departure. It must pause first, however, to deal with Spiegelman's novel argument that the Court may not depart in this case, no matter what the facts, because it is bound by the plea agreement between Spiegelman and the government — most notably the sentencing stipulations and the agreement that neither the government nor the defendant would seek a departure.

As a general matter, of course, it is well settled that district courts may depart from the Sentencing Guidelines sua sponte. Moreover, Spiegelman's argument concerning the sentencing stipulations is easily disposed of. Section 6B1.4 of the Guidelines provides that "[t]he court is not bound by the [sentencing] stipulations, but may with the aid of the presentence report, determine the facts relevant to sentencing." The official commentary eliminates whatever doubt there might be:

"The court is not obliged to accept the stipulation of the parties. Even though stipulations are expected to be accurate and complete, the court cannot rely exclusively upon stipulations in ascertaining the factors relevant to the determination of sentence. Rather, in determining the factual basis for the sentence, the court will consider the stipulation, together with the results of the presentence investigation, and any other relevant information."

Nor does the parties' agreement that neither would seek a departure constrain the Court. As Spiegelman was warned before he entered his guilty plea, their agreement by its terms binds only the parties. If it purported to bind the sentencing judge, moreover, it would be of no effect. Section 3553(b) of Title 18 provides that the sentence imposed must be of the kind and within the range prescribed by the Sentencing Guidelines "unless the court finds that there exists an aggravating or mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission in formulating the guidelines that should result in a sentence different from that prescribed." There is no suggestion that the Court may be divested of that authority by agreement of the parties, a view confirmed by the introductory commentary to the portion of the Guidelines dealing with plea agreements, which specifically states "that sentencing is a judicial function and the appropriate sentence in a guilty plea case is to be determined by the judge." Further, Section 5K2.0 of the Guidelines specifically provides that "[t]he controlling decision as to whether and to what extent departure is warranted can only be made by the courts."

Undaunted by the plain terms of the governing statute and the Sentencing Guidelines, Spiegelman argues that the Court nevertheless was stripped by the plea agreement of its statutory responsibility to determine whether a departure is appropriate by United States v.. Torres-Echavarria.

The Second Circuit in Torres-Echavarria affirmed a distr ct court's rejection of an agreement pursuant to which the defendant agreed to plead guilty to a lesser charge in exchange for the government's agreement to dismiss a more serious accusation. Although the Circuit acknowledged that the "plea bargain is an indispensable tool for the administration of the criminal law," it reasoned that a court has "broad latitude" to accept or reject a proposed plea agreement that involves a bargained-for dismissal of one or more charges in return for the defendant's plea to a less serious offense because the "review of a prosecutor's proposal to dismiss a claim . . ., in consideration of a plea of guilty to some other offense . . ., implicates core judicial functions [sentencing]."

Spiegelman, making selective reference to Torres-Echavarria and the commentary to U.S.S.G. § 6B1.2, argues that:

"[I]n situations involving a plea agreement calling for dismissal of some charges, the sentencing court `should defer to the government's position except under extraordinary circumstances.' . . . . It follows that when the plea agreement (as in Spiegelman) requires a plea to the entire information, we no longer have an `extraordinary circumstance', and there just is no support in the guidelines or the case law for the sentencing court to reject (or not fully implement) the agreement."

The argument is without merit.

The Court of Appeals made two points in Torres-Echavarria. First, district courts ordinarily should defer to the government when it decides to drop charges unconditionally because the charging decision is at the heart of the prosecutor's function. On the other hand, the same deference would be inappropriate when a decision to dismiss is contingent upon the defendant's agreement to plead to a remaining charge. In that situation, "the court's authority to adjudicate guilt and impose sentence is implicated." The court therefore must ensure that the remaining charge adequately reflects the seriousness of the offense and that an appropriate sentence can be imposed. In other words, Torres-Echavarria illustrates that the Guidelines regime functions to strengthen the discretion of both prosecutors and judges within their respective realms. For prosecutors, that realm includes the charging decision. Sentencing, however, is the exclusive province of the judiciary, within the constraints of the Guidelines and the Sentencing Reform Act.

In this case, Spiegelman pleaded guilty to all of the charges against him. The plea agreement involved no exercise of the United States Attorney's charging discretion. This Court had no discretion as to whether to accept the plea. Its sole function is to determine an appropriate sentence, which 18 U.S.C. § 3553(b) and the Sentencing Guidelines make clear is purely a judicial task. Nothing in Torres-Echavarria is to the contrary. Indeed, its reiteration that district courts have "broad latitude" to reject plea agreements that implicate their sentencing function supports the conclusion that the parties' agreement in this case that neither would seek a departure has no effect on the Court's ability to consider departing.

IV.

Departures Under the Guidelines

Section 3553(b) of Title 18, which is implemented by Section 5K2.0 of the Guidelines, permits a district court to depart from the Guideline range if it "find[s] that there exists an aggravating or mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission in formulating the guidelines." Departures are appropriate where the Guidelines do "not adequately take into account cases that are, for one reason or another, `unusual.'"

The Guidelines "list certain factors which never can be bases for departure. . . . but then state that, with the exception of those listed factors, [they] `do not intend to limit the kinds of factors, whether or not mentioned anywhere else in the guidelines, that could constitute grounds for departure in an unusual case.'" The Guidelines then go on to provide "considerable guidance as to the factors that are apt or not apt to make a case atypical, by listing certain factors as either encouraged or discouraged bases for departure." If a factor is not mentioned in the Guidelines at all — that is, it is neither a forbidden nor an encouraged basis for departure — "the court must, after considering the `structure and theory of both relevant individual guidelines and the Guidelines taken as a whole,' decide whether it is sufficient to take the case out of the Guideline's heartland." Section 5K2.5

Section 5K2.5 provides:

"If the offense caused property damage or loss not taken into account within the guidelines, the court may increase the sentence above the authorized guideline range. The extent of the increase ordinarily should depend on the extent to which the harm was intended or knowingly risked and on the extent to which the harm to property is more serious than other harm caused or risked by the conduct relevant to the offense of conviction."

Here, the principal guideline applicable to Spiegelman's offense is Section 2B1.1 which, like other guidelines applicable to property crimes, provides for a base offense level where the loss is small and for increases in the offense level according to the dollar amount of the loss. As the sole measure of loss used in Section 2B1.1 is the economic value of the property stolen, Section 2B1.1 does not adequately measure the damage caused by a theft such as this one, which risked and caused loss that is not fully measurable in economic terms. Hence, Section 5K2.5 appears to be directly on point and to encourage a departure here.

In an argument first raised only days ago, despite the pendency of this matter for nearly a year, Spiegelman contends that Section 5K2.5 has no application here because it encourages departures only in cases in which the defendant causes or knowingly risks property damage or property loss not adequately taking into account by the Guidelines — not property damage or loss of the sort at issue here. He reaches this conclusion on the theory that the second sentence of Section 5K2.5 makes the extent of any upward departure dependent upon "the extent to which the harm to property is more serious than other harm caused." He relies also on the Ninth Circuit's decision in United States v. Dayea.

Spiegelman's argument is unpersuasive. He doubtless is correct that rules of grammar suggest that the phrase "property damage or loss" means "property damage or property loss." But that does not necessarily mean that the phrase "property damage or property loss," in an appropriate case, cannot comprehend the proximate consequences of damage to or loss of property. For that proposition, he posits that the second sentence of Section 5K2.5 provides that "[t]he extent of the increase . . . depend[s] on the extent to which harm to property is more serious than other harm caused . . . by the offense of conviction." He then goes on to argue that Section 5K2.5 requires weighing harm to property against other harm caused by the offense of conviction, "thereby completely precluding any possibility of applying this section to an offense against property."

Among the problems with this argument is that it rests on Spiegelman's misquotation of the second sentence of Section 5K2.5. The sentence actually provides that "[t]he extent of the increase ordinarily should depend on the extent to which the harm was intended or knowingly risked and on the extent to which the harm to property is more serious than other harm caused or risked by the conduct relevant to the offense of conviction." This is not surprising. In most circumstances, Section 5K2.5 will be relevant to non-property crimes in which the defendant causes serious harm to property as well. As guidelines levels for such crimes are set without regard to any harm to property and often are higher than those for property crimes, the second sentence of Section 5K2.5 indicates what is relatively obvious in any case: that "ordinarily" any harm to property should be treated as an aggravating factor in such cases only where that harm materially changes the overall gravity of an already serious offense. Similarly, Section 5K2.5 usually will be irrelevant to property crimes because the value of the property stolen or destroyed typically is a perfectly adequate measure of the gravity of the offense, as the only harm caused by the offense of conviction usually will be the direct economic loss. But the fact Section 5K2.5 "ordinarily" will not be relevant to property crimes certainly does not establish that it never will be relevant. It does no violence to the second sentence of Section 5K2.5, which in any case is merely a hortatory policy statement rather than an imperative, to recognize that where a property crime causes serious non-economic loss, certainly an extra-"ordinary" situation, the non-economic loss may be taken into account along with the monetary loss in considering a departure. In the last analysis, however, this debate about Section 5K2.5 is immaterial because a departure is appropriate under Section 5K2.0 in any event.

Section 5K2.0

Section 5K2.0 permits a departure whenever "there exists an aggravating or mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission in formulating the guidelines." The non-monetary loss risked and caused by Spiegelman is such an aggravating factor, as Section 2B1.1 measures the gravity of theft offenses solely in terms of monetary loss. Hence, an upward departure is appropriate under Section 5K2.0. Indeed, the only case directly in point supports that conclusion.

This view is reinforced by analogy to other sections of the Guidelines which employ the same approach to determining the gravity of property crimes. Application Note 10 to Section 2F1.1, the fraud guideline, for example, provides that an upward departure may be warranted where "the fraud caused or risked reasonably foreseeable, substantial non-monetary harm." Application Note 4 to Section 2B1.3, which applies to property damage and destruction, is to similar effect. And these analogies are especially persuasive because Sections 2B1.3 and 2F1.1 define "loss" by reference to Section 2B1.1. There is no principled basis for reaching a different result where non-monetary harm is caused by theft.

The plain language of Sections 5K2.0 and 2B1.1 and the application notes to Sections 2B1.3 and 2F1.1 thus support the conclusion that an upward departure is appropriate in this case in view of the Court's findings that Spiegelman's theft both knowingly risked and caused substantial non-monetary damage to Columbia, to scholars in general, and to society. And if there were any doubt on that score, it would have been eliminated by a recent amendment to the Guidelines Manual.

The Sentencing Commission, effective November 1, 1997, amended the Guidelines Manual to add a new Application Note 15 to Section 2B1.1, which states:

"In cases where the loss determined under subsection (b)(1) does not fully capture the harmfulness of the conduct, an upward departure may be warranted. For example, the theft of personal information or writings ( e.g., medical records, educational records, a diary) may involve a substantial invasion of a privacy interest that would not be addressed by the monetary loss provisions of subsection (b)(1)."

Application Note 15 thus reinforces the conclusion the Court would reach in any case because it confirms that the Court may depart upward when the economic loss does not account for all aspects of the harm caused by the underlying offense. As the Eleventh Circuit has held, clarifying amendments to the Guidelines commentaries "constitute strong persuasive evidence of how the Sentencing Commission originally envisioned that the courts would apply the affected guideline."

There is another point demonstrating that an upward departure is appropriate here, although what has been said to this point is more than sufficient. Section 2B1.1 employs economic loss as a measure of the seriousness of the crime because the Sentencing Commission regards the economic loss "as an indicator of both the harm to the victim and the gain to the defendant." It takes no account of harm to others. Spiegelman's offense, however, caused harm not only to Columbia, his immediate victim, but to a much broader community. Hence, as the Second Circuit's recent decision in United States v. Malpeso indicates, a departure is appropriate in this case because Section 2B1.1 does not adequately consider the harm knowingly risked and inflicted by Spiegelman on persons other than Columbia University.

Spiegelman nevertheless argues that there is nothing about this case that takes it out of the "heartland" of Section 2B1.1 because other thefts of rare books and manuscripts have not resulted in departures or have been treated leniently. He points to three cases involving thefts from the Library of Congress and a 1992 state court disposition of a case involving a theft of rare documents from Columbia. The argument is entirely unpersuasive.

Spiegelman has provided the Court with substantially no information concerning the nature and extent of the thefts at issue in the three District of Columbia cases. There is no basis for concluding that the scale, impact or nature of those crimes even remotely approached Spiegelman's offense. Nor is there any reason to suppose that the sentencing judges in those cases had before them evidence of the character present here. While the theft at issue in the state court case appears to have been substantial, the Court does not regard the disposition as having any material bearing here because the case was not governed by the Sentencing Reform Act and the Sentencing Guidelines. Indeed, the sentence in that case was far below the Guideline range that would have applied in federal court based purely on the economic loss. In any event, there is no denying that the "heartland" of Section 2B1.1 is garden variety theft — theft of money or fungible property — which causes only economic harm and impacts only the immediate victim. This case manifestly is exceptional by virtue of the substantial and unquantifiable non-economic harm risked and caused both to Columbia and others. It is outside the heartland almost by definition.

Spiegelman points also to a lengthy list of thefts of rare books and manuscripts and asserts that a departure here would be inappropriate because there were no departures in any of those cases. But the list for the most part does not even disclose whether the guilty parties were apprehended, much less whether the cases were prosecuted in federal courts or resulted in convictions. Nor does the Court accept counsel's unsubstantiated assertion that there were no departures. The list, however, is important in one respect. It demonstrates that theft of rare books and manuscripts from academic institutions is rampant, and it emphasizes the need for better deterrence.

The fact that some of the stolen property was recovered as the result of information provided by Spiegelman after his arrest does not alter the conclusion that a departure is appropriate. Among the fundamental premises of the Sentencing Guidelines is that "loss" is measured with reference to the property stolen or otherwise wrongfully acquired, irrespective of whether the property later is recovered. Accordingly, the Court considers the recovery of some of the property as bearing only on the extent of any departure, not on the question whether a departure is appropriate.

For the foregoing reasons, this Court holds that the non-monetary losses knowingly risked by Spiegelman and inflicted by him on Columbia and on scholars and the nation at large are an "aggravating . . . circumstance of a kind . . . not adequately taken into consideration by the Sentencing Commission in formulating the guidelines." It therefore elects to depart upward in order to fix an appropriate sentence.

V.

Spiegelman intentionally or knowingly risked inflicting, and inflicted, substantial harm not only upon his immediate victims, Columbia University and its professors and students, but also upon the greater academic community and society as a whole. In callously stealing, mutilating, and destroying rare and unique elements of our common intellectual heritage, Spiegelman did not simply aim to divest Columbia of $1.3 million worth of physical property. He risked stunting, and probably stunted, the growth of human knowledge to the detriment of us all. By the very nature of the crime, it is impossible to know exactly what damage he has done. But this much is clear: this crime was quite different from the theft of cash equal to the appraised value of the materials stolen because it deprived not only Columbia, but the world of irreplaceable pieces of the past and the benefits of future scholarship. To treat Spiegelman's offense as being of the same gravity as the theft of $1.3 million in cash would be to deny the unmistakable importance of the undiscovered knowledge likely buried within the items he stole. Thanks to Spiegelman, that knowledge now may remain forever undiscovered.

Congress has instructed courts in fixing sentences to consider, among other things, the nature and circumstances of the offense, the need to deter others and protect the public from similar crimes, and the need to reflect the seriousness of the offense. If the economic loss were the sole measure relevant here, Spiegelman would be facing 30 to 37 months imprisonment. The statutory maximum for the most serious offense of conviction is 10 years. An upward departure of five levels, from level 18 to level 23, would bring the Guideline range to 51 to 63 months, which is somewhat less than double the range reached on the basis of the economic loss alone but far less than the statutory maximum. It is a range which, in the Court's judgment, is necessary to make clear both to Spiegelman and to others considering similar conduct that theft of irreplaceable intellectual and cultural resources is a very serious matter and will not be dealt with purely in economic terms that disregard important aspects of the harm that it causes.

Accordingly, this Court departs upward from the Guideline range by five levels resulting in an adjusted Guideline range of 51 to 63 months.

The foregoing constitute the Court's findings of fact and conclusions of law.

SO ORDERED.