Weapons labs and the inconvenient truth

On February 16, the strategic forces subcommittee of the House Armed Services Committee held a hearing on the oversight of the nuclear weapons laboratories. The nine out of 16 committee members who skipped the hearing — including the congressman who represents the district encompassing the Livermore Laboratory — missed a fine show.

Worth the price of admission alone was Paul Robinson, the former director of the Sandia National Laboratories, who has a reputation for wandering off script in congressional testimony. He did not disappoint. Giving a rambling monologue in the style of an indulged grandfather sitting on the family porch, he suggested that many of the labs’ problems could be solved if they were transferred from the Energy Department to the Defense Department, just like the good old days of the Manhattan Project. Maybe he wasn’t paying attention to the precipitous decline in morale at Los Alamos when the United States put it under the command of an admiral, Pete Nanos, from 2003 to 2005. And if, as Robinson argued, the labs’ main problem was excessive bureaucracy, handing them over to the agency that inspired Catch-22 is an odd remedy.

But the star attraction was a bravura act of logic-juggling by Charles Shank and Charles Curtis, two of the authors of a new National Academy of Sciences report, Managing for High-Quality Science and Engineering at the NNSA National Security Laboratories. Bolstered by testimony at the hearing from three emeritus directors of Los Alamos, Livermore, and Sandia, Shank and Curtis argued that employee morale was nose-diving and that the labs were in danger of losing their best scientists — because the Energy Department’s National Nuclear Security Administration has been micromanaging the labs and burying scientists under unreasonable bureaucratic requirements that obstruct scientific work. The National Academy of Sciences report cites an example of a scientist who complained that he needed five signatures from managers to attend a physics conference. Panelists also suggested that cutting Energy Department red tape would save money better spent on scientific research.

And then something strange happened: A few minutes later in the hearing, Shank and Curtis denied suggestions that morale at the Livermore and Los Alamos labs had been harmed by the shift in management from the University of California to various limited liability consortia, both led by Bechtel. Their reasoning: The evidence in support of this claim was merely anecdotal, but the actual numbers show that no more scientists are leaving the labs now than was the case a few years ago; so, clearly, the labs’ employees are not unhappy.

Morale is nose-diving, but the scientists are not unhappy?

Shank is an engineer and Curtis is a lawyer, so maybe neither took any logic classes. But surely there is a problem here that should be obvious to any undergraduate philosophy major. Let’s spell it out.

Four propositions were at issue in the hearing:

• Proposition 1: Energy’s micromanagement and bureaucracy is damaging scientists’ morale at the labs and threatening employee recruitment and retention.

The panelists agreed.

• Proposition 2: More money would be available for the labs’ scientific mission if Energy’s bureaucratic oversight of the labs were cut.

The panelists agreed.

• Proposition 3: The recent privatization of the labs under Bechtel’s lead is damaging scientists’ morale at the labs and threatening employee recruitment and retention.

The panelists disagreed.

• Proposition 4: More money would be available for the labs’ scientific mission if they were not managed by the Bechtel consortia.

The panelists disagreed. (When asked how to save money to increase the labs’ science budgets, no one suggested putting the labs back under university management. So this will count as a disagreement with the proposition.)

There is an obvious logical contradiction between the panelists’ positions on proposition 1 and proposition 3. Shank told the congressmen that any evidence that Bechtel’s management had harmed employee morale was unreliable because it was entirely anecdotal. But anyone who reads the National Academy of Sciences report will find that the evidence of Energy’s over-management harming employee morale is, well, anecdotal. I sometimes joke that, for anthropologists like me, data is the plural of anecdote, so anecdotal evidence does not bother me, as long as it is systematically gathered and evaluated. But what is not okay is to rule personal testimony out of bounds in one part of the report and then make it the bedrock of your case in another. Aristotle would disapprove.

Furthermore, if it is the case that employee attrition rates have been stable for years, this calls into question the report’s claims that excessive bureaucracy has damaged morale at the weapons labs as much as it does similar claims about Bechtel’s management style. In regard to morale, propositions 1 and 3 sink or swim together.

Of course, crude numerical rates of employee attrition may not offer the best measure of employee morale — especially in the midst of the worst recession in living memory. It is possible that many weapons scientists are itching to leave Livermore and Los Alamos, but cannot do so in this job market. One senior manager at Livermore suggested to me that Bechtel’s management style — which tends to combine the worst aspects of the Department of Motor Vehicles and Goldman Sachs — had broken the dedication of many scientists and, while they did not leave, nor did they feel the need to work long hours. “I don’t see many lights on at night any more,” he said. It would be interesting to survey scientists on overtime worked since Bechtel took the reins. And according to The New York Times‘ Bay Area edition, the number of peer-reviewed articles published at Livermore declined from 1,400 in 2005, before Bechtel took the helm, to 800 in 2010 under Bechtel. That doesn’t sound like an organization with stable morale.

Finally, there is the issue of budgetary savings. Pressed to specify how taxpayers’ money might be saved and reallocated to scientific work, the panelists’ only suggestion was to cut bureaucratic red tape and reduce the number of Energy oversight officials. (That is, they endorsed proposition 2, but ignored proposition 4.)

But what about cutting the costs associated with the new management contract? After all, we learn from the National Academy of Sciences report itself — page 11 — that the management fee for Los Alamos increased from around $10 million a year to $60 million and, for Livermore, from around $10 million to $40 million when the contract passed from the University of California to the different Bechtel-led consortia. That is an increase of $80 million a year. At the same time, since managerial responsibility shifted from a nonprofit (the University of California) to a for-profit (the limited liability corporations), the labs suddenly became responsible for tens of millions of dollars in taxes. And, because they withdrew from the university system’s massive employee pool, the labs incurred higher benefit costs. Mike Anastasio, former director of Los Alamos, put the cost of the new taxes and benefits alone at his lab at “$225 million per year.” Add in the increased management fee, and you get to $275 million. Yet no one told the assembled congressmen that going back to management by the University of California could free up $275 million, which then could be reallocated to actual science.

The National Academy of Sciences report does, however, mention that the extra $80 million in management fees “is a small fraction of the total operating budget of the Labs” — page 12 — and is thus of little consequence. It’s funny how tens of millions of dollars seem like chump change to those who do government studies when it goes to private contractors. I doubt the Tea Party would see things this way.

As an anthropologist who studies the culture of the weapons labs, I think the panel was absolutely correct that Energy micromanagement is harming the labs. But the National Academy of Sciences report engages in pretzel logic by giving Bechtel and the new management consortia an undeserved free pass — so much so that one has to wonder if the report was cooked. To be sure, there were problems with accounting and other kinds of managerial oversight under the University of California, but I know from conversations and correspondence with many lab employees (more anecdotes) that the management style of the new Bechtel consortia is deeply resented and has depressed employee morale. That style consists of rotating managers in and out from corporate headquarters, rewarding paperwork measures of safety over scientific productivity, widening pay differentials, managerial obsession with bonuses, corporate information control in place of a culture of academic freedom, and, in the case of Livermore, massive layoffs executed with the finesse of Donald Trump.

You will learn none of these things about the new management from the National Academy of Sciences report. The report fits nicely with currently fashionable narratives about the evils of government bureaucracy and the redemptive power of private contractors. Unfortunately, it doesn’t fit so well with the facts.