The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 56.0°F | A Few Clouds

Provost Denies Postol’s Request for ABM Review

By Keith J. Winstein


Provost Robert A. Brown has denied a request by Professor Theodore Postol to appoint an independent committee to review a controversial 1999 missile defense report, calling it a “government, not MIT, document.”

“MIT Lincoln Lab had the leadership role” in the report, said a source close to the team which produced it.

Postol, professor of science, technology and national security policy, had first asked for MIT to review the disputed document in April 2001.

“I believe that MIT has an obligation to the country to independently review this study and pass judgment on whether it meets scientific standards of our institution,” wrote Postol in an April 26, 2001 letter to President Charles M. Vest.

Brown’s denial of Postol’s request focused on whether MIT has a responsibility to pass judgement on the document, which was commissioned in 1998 by the Ballistic Missile Defense Organization (BMDO) to independently review allegations of fraud in a June 1997 test for the Pentagon’s missile defense system.

“To the extent that you seek a review of the factual accuracy of the conclusions of the [study], which is a government, not MIT, document, there is no basis for MIT to act,” Brown wrote in a Feb. 11 letter to Postol. “Although you may not agree, I believe that MIT as an institution should not -- and in fact, does not -- put itself in the position of being an arbiter of the accuracy of scientific reports issued by other institutions, governmental agencies, or others.”

Review started regarding authors

The missile defense report is entitled “POET Study 1998-5: Independent Review of TRW Discrimination Techniques Final Report,” and bears MIT’s name only in its listing of authors.

Two of the report’s five authors are senior staff members at MIT’s Lincoln Laboratory, and their names precede the text “MIT Lincoln Laboratory” on the report’s cover sheet.

Although Brown declined Postol’s request for MIT to review the document itself, he wrote that he would “initiate a review of [Postol’s] allegation ... pursuant to Section 10.1 of MIT’s policies [regarding academic misconduct]” into whether the two MIT co-authors had committed scientific misconduct in their work for the team, something Postol had not asked for.

“I’ve never given the names” of the MIT authors, Postol said. “I brought this to Vest last April privately. I have no desire to see anybody punished.”

Administration sources stressed that a review of allegations under MIT’s academic misconduct policies does not mean that there is an actual investigation of misconduct.

“The mere fact that an allegation of academic misconduct has been made is sufficient to trigger an inquiry,” Brown wrote. “The initiation of an inquiry should not be taken as an indication that MIT has determined that the persons against whom the allegation has been made has done anything improper. The purpose of the inquiry is to ascertain if there is sufficient substance to the allegation to warrant the initiation of the investigation.”

Brown was unavailable to comment for this article.

Charlene M. Placido, the assistant dean for research, said, “Our scientific misconduct cases are completely confidential ... No one has ever known publicly about any of our cases.”

In addition to the repeated requests for an independent review of the report, Postol wrote to MIT Corporation Chairman Alexander V. d’Arbeloff ’42 on Jan. 14, 2002 “to lodge a formal complaint against President Charles M. Vest for failing to investigate a serious case of scientific fraud that has taken place under his oversight of the MIT Administration.”

Kenneth D. Campbell, director of the News Office, said the complaint will be handled in accordance with Section 9.6 (“Complaint and Grievance Procedures”) of the MIT Policies and Procedures.

Team studied decoy identification

The disputed document was authored by the BMDO’s Phase One Engineering Team, known as POET. BMDO, once known as the Strategic Defense Initiative, was renamed the Missile Defense Agency in January 2002.

The BMDO commissioned the study to review allegations of fraud in the June 1997 Integrated Flight Test 1A, which had sought to examine whether the techniques then in development by defense contractors Boeing Company and TRW Inc. could successfully distinguish between an attacking nation’s missiles and the balloon decoys that it could use to confuse a missile defense system.

The study concluded that, contrary to allegations of fraud in the test, the Boeing-TRW system “functioned as designed,” said Lt. Col. Rick Lehner, spokesman for the agency.

“We’re done talking about [the 1999 study],” Lehner said, noting that an FBI review of the report concluded that there had been no evidence of fraud.

Lehner said the widely-reported disputes over the quality of the report are irrelevant. The Boeing-TRW system then at issue “was not selected for missile defense,” he said, the agency having instead favored a system developed by Lexington-based Raytheon Company.

Source disputes Postol allegations

Even if MIT did commission an independent review panel of the document as Postol had requested, there is of course no guarantee that it would come to the same conclusions as Postol.

“Ted Postol is making all these claims,” said the source close to the POET team. “I would say he doesn’t know what he’s talking about.”

“Unfortunately the team can’t talk to him and explain to him that it has information that would disprove his allegations,” the source said.

The team “was told not to talk to him,” said the source, who would only say that “officials from Washington” had given the instructions.

The source said that “the POET team took extra precaution to be able to work independently and approached the controversy from an unbiased viewpoint,” disputing Postol’s assertions that the team had been manipulated by the BMDO.

The source flatly denied Postol’s assertion that arithmetic mistakes and a problematic statistical methodology had caused errors in the report’s “confusion matrix,” which represents the probability that various airborne objects, including warheads and balloon decoys, can be mistaken for each other.

As the matrix is printed in the report, a tumbling warhead is 40 times more likely to look like a particular kind of balloon than to be correctly identified by a missile defense sensor. Postol had pointed out this confusing discrepancy in a March 2001 letter to the General Accounting Office, saying that it revealed that “the most cursory of ‘sanity checks’ would have revealed to the POET analysts that something was seriously wrong with their conclusions.”

In January of this year, Postol performed an analysis on the single graph of raw data included in the report, concluding that the level of noise present indicated that the sensor had failed to cool to its operating temperature and had thus “provided no usable data” during the period of the test when it needed to try to distinguish warheads from decoys.

Discussions of the sensor’s failure to cool were included in the test’s “60-day report,” Postol said, and were reported in Science magazine on Feb. 1, 2002, but were not mentioned in the POET report.

Postol asked, “How is it possible that MIT Lincoln Lab would know that the sensor failed to perform and they would allow their people to publish a report as if the data were valid?”

The source close to the team disputed Postol’s allegations that temperature problems had caused the sensor to return no usable data. “That doesn’t mean that the sensor doesn’t work,” said the source. “The flight experiment yielded useful data for our analysis. The results were also corroborated with measurements from other independent off board sensors .... The statement [that the sensor returned no usable data] is incorrect. To find the signal, one has to know where to look.”

“In my opinion, Professor Postol is using the good name of MIT and promoting his own agenda,” said the source. Referring to work in the early 1990s when Postol analyzed news footage to dispute government assertions regarding the Patriot missile, the source said, “I respected his study on the performance on the Patriot missiles, but he has made several serious technical blunders in his evaluation of the POET report.”

Postol is candid that his understanding of what he calls the confusing report is not perfect. “I think I might have said that at one point and I think I was wrong,” he said in response to a question about one critique of the report he had sent to the White House. “I want to be very clear: my understanding of this has evolved.”

Report alleged to halt DoJ inquiry

Postol said it is critical that MIT examine the study because “this is a matter of national import.” The Pentagon, Postol said, is “misleading the Congress with the help of the MIT Lincoln Laboratory study.”

According to Postol, the document was partially responsible for the decision of the Department of Justice and the Defense Criminal Investigative Service to halt investigations into the fraud allegations that had surrounded the 1997 test.

Postol believes the allegations, levied by former TRW engineer Nira Schwartz in a federal lawsuit, have merit.

The BMDO “used the imprimatur of MIT” to stop the investigations, Postol said.

Robert Blumenfield, a district director for Congressman Howard L. Berman of California, confirms that the POET report was “certainly one of the factors” leading the Justice Department to terminate an investigation into the fraud allegations.

Blumenfield said the General Accounting Office, the nonpartisan investigatory arm of Congress, is investigating the reasons for the Justice Department’s decision at Berman’s request.

“POET’s data and analysis of the data could not possibly justify their conclusions,” Postol wrote to the General Accounting Office in March 2001. The study “was by design created to make the false impression that something was true when it was not. To put it more succinctly, the document is fraudulent.”

MIT should review the report, Postol said, and “if it does not prove what it claims to do, all they need to do is write a letter to the Department of Defense and the Department of Justice,” reiterating in a Feb. 12 letter to d’Arbeloff that “I also reject the claim that it is necessary to go after the MIT Lincoln Laboratory authors of the document.”

Dispute over MIT’s responsibility

What is currently at issue is whether MIT is responsible for the quality of the study. In nine letters to senior members of the MIT administration since April 2001, Postol had repeatedly asked MIT to commission a review of what he at one point called “the fraudulent scientific document” that was “produced by the MIT Lincoln Laboratory.”

Postol’s scientific concerns with the report have been widely reported, but his requests for MIT to initiate an independent review were first disclosed in a Boston Globe article published on Feb. 9, 2002.

Provost Brown’s decision was criticized in a Feb. 15, 2002 editorial in the Globe, which asked “why administrators at [MIT] have been exceedingly slow to respond to the repeated warnings of [Postol] that MIT has been implicated in what he believes is scientific fraud.”

“The school’s administration ought to reconsider this self-protecting institutional reflex” not to commission an independent review of the report, the Globe wrote.

Several MIT officials gave varying explanations of the criteria used by Brown to decide that the POET study is a “government, not MIT, document” in his letter to Postol.

In his letter, Brown referred questions to Placido, the assistant dean for research. Placido initially said that she didn’t believe MIT research sponsored by the government “would be a government document because all research belongs to MIT even if it’s federally sponsored.”

However, Placido then clarified that she was referring to whether a report was a “government document” for purposes of patent ownership, something which she said was not necessarily related to whether research is a “government document” for purposes of an MIT scientific misconduct investigation.

However, Campbell, of the MIT News Office, drew a link between patent ownership and the criteria which constitute a “government document” to decide whether MIT should commence a review.

“Government documents are not documents that deal with licensable or copyrightable intellectual property,” Campbell said. “This is an analysis for the government by individual scientists belonging to three different institutions and it is not an MIT document.”

“As an institution, we’re not in the business of saying, well, that government report is wrong, and we’re going to investigate it. That’s not our role,” Campbell said.

A source close to the administration said that MIT’s only method for passing judgment on the report would be through a misconduct investigation targeted at the particular authors.

Under the MIT Policies and Procedures, the source said, “there’s no connection to a government document for purposes of a misconduct investigation except as MIT individuals were involved in the report.”

MDA contract is $100M/year

Roger Sudbury, assistant to the director of Lincoln Laboratory, said that “the POET report was a national study that included two Lincoln Lab staff members on the team, and one of the Lincoln Lab members was chair of the study team.”

“It was a POET document,” not an MIT document, Sudbury said. “It was part of a larger activity and our POET support.”

“All of our missile defense activities result in funding of about $100 million a year,” Sudbury said, but that includes “a large array of activities.”

“We wouldn’t have a breakout” in the budget on the particular efforts which produced the report in question, Sudbury said. “It was not a separate item.”

“We don’t look at [sponsored research] as a government document or an MIT document,” said Paul C. Powell, assistant director of the Office of Sponsored Programs. “It’s just a contract between us and the federal government.”

The contractual relationship, Postol said, is one reason MIT should be responsible for the quality of the document’s research.

“It may well be a government document in the end. If MIT was responsible for the accuracy of the document by contract, then MIT is responsible for correcting the multiplicity of errors in the document, if MIT knows the document is in error,” Postol said.

The Boston Globe reported in its Feb. 9 article that the report has “won respect in some Washington missile defense circles as the ‘MIT study’.”

Postol’s description of a July 2001 conversation with Jeffrey Swope, MIT’s primary outside counsel, appeared to contain a different MIT view regarding what constitutes an “MIT document.”

In the widely-reported incident, the Defense Security Service asked MIT on July 10, 2001 to confiscate documents from Postol’s office, including correspondence Postol had written to the General Accounting Office. MIT eventually decided not to comply with the government’s request to confiscate the documents.

Postol said in a recent interview that Swope had attempted to justify confiscation of Postol’s letters in a July conversation on the basis that “the argument can be made that something which goes out with MIT’s name is an MIT document.”

Swope declined to comment on the conversation.

GAO reviewing Postol allegations

Postol’s allegations are the subject of two separate reviews by the General Accounting Office, requested by Congressman Berman and Congressman Edward J. Markey of Massachusetts. The Berman study is mostly based on the allegations of Schwartz, the former TRW engineer, while the Markey study has focused more on Postol’s allegations, Blumenfield said.

The reviews will probably be published “within a month,” Blumenfield said.

Postol has been asked to discuss the investigations or drafts of the GAO reviews with the offices of both Congressmen, staffers confirmed.

Postol declined to comment on the investigations, despite his involvement. “It’s proper for them to release the information, not me.”

“I think the reports will significantly advance our understanding of the situation,” he said.