Diamonds, Water, and Sandy Berger

Don't be fooled. Sandy Berger did not make a careless mistake. The claim that he did is based on the careful use of one side (the incorrect one) of what is known in economics as the diamonds and water paradox.

For those of you are not following this news story, Sandy Berger was an advisor to the Kerry campaign (he resigned on July 20th), and National Security Advisor during the Clinton administration. He has admitted to taking classified documents regarding America's terrorist preparations from the National Archives.

A lot of weak thinking is being passed about regarding the Sandy Berger situation. The marginal analysis that economists are trained to do can illuminate what's going on here.

The weak thinking is made by apologists for Berger who note that the potential personal cost to him of taking those documents far outweighs any conceivable benefit that he could have derived from them. Therefore, they conclude that this must have been a careless mistake.

This is a story that doesn't raise any suspicions, except for that odd fact that he apparently hid some documents in his socks.

So, let me lead you through this story using the methods that economists use to think about how people make decisions.

First off, the argument being offered by Berger's supporters is an example of the water and diamonds paradox. This is the idea that it is strange that diamonds cost more than water even though the latter is more valuable to us (since we'll die without it). The paradox is reconciled by recognizing that the prices we are willing to pay don't reflect the total benefit we get from goods. Rather they reflect the marginal benefit we get from each unit of the good that we buy. If you had no water and no diamonds (and were obviously dying of thirst too), which would you pay more for? Water is the obvious answer. But, you and I are rarely in that bizarre position, so why is water generally cheaper than diamonds? The answer is that we are comparing the price  of the most recent of many purchases of water to the rare and perhaps only purchase of diamonds that we ever make. The conclusion is that all of our water would cost more than all of our diamonds if we ever had to make that purchase, but our last (marginal) purchase of water costs less than our last (marginal) purchase of diamonds.

If you buy into the argument that Berger made a careless mistake since the total cost of taking classified documents was less than the total benefit he could earn by taking them, then you are being victimized by the skillful use of the paradoxical part of the water and diamonds allegory. Mark my words, it is smoke and mirrors, and many people are falling for it.

What we need to do is analyze the situation from the perspective of marginal costs and marginal benefits. So, what are the facts?

1) Berger went to the National Archives to do research for his testimony before the 9/11 commission.
2) Berger viewed classified documents.
3) Berger did not send an assistant to do this work.
4) Berger removed classified documents from the National Archives.
5) Berger hid his actions.
6) When confronted the next day, he admitted to taking the documents, and willingly returned them.
7) Berger removed some documents again after being caught the first time!
8) Berger never kept any documents for good (although this is still not clear).

What does marginal analysis tell us about these events?

1) Berger had something to gain by examining the documents. This is perfectly sensible. (But, I am curious how many other Clinton and Bush administration officials had to go review documents related to their own past actions. If you find that degree of forgetfulness bothersome, then I think there is a useful Freedom of Information request there for someone more ambitious than me). So, the marginal benefit of doing the research exceeded the marginal cost.
2) What does classified mean? It is the government's way of saying you have to pay a high price to view the document. This can be because it was expensive to create, or extremely useful to those who might read it. Since we're not talking about a patent here, it is pretty clear that Berger went to view documents that others thought were extremely valuable. He apparently did too. Since prices reflect marginal valuations, the classified status of the documents indicates that they were of high marginal benefit, that certainly exceeded the marginal cost of reading them.
3) Why didn't Berger have an assistant do the work? The obvious answer is because the documents were classified. This doesn't really hold water though, since there are lots of people in D.C. with security clearances for hire. And, people in Berger's position tend to have assistants for just such work (in fact, I'd bet the 9/11 commission would have even paid for a research assistant). The implication here is that Berger could have gotten the marginal benefit of the research by having someone else do it, but that the marginal costs of having someone else do it were too high, so he did it himself.
4) Berger removed the documents from the National Archives. The penalties for this are stiff, so the marginal cost of this act was high. But the marginal benefits must have been even higher or he wouldn't have done it.
5) Berger understood that the marginal costs of taking the documents was high, and chose to avoid them by hiding the documents in his personal effects (i.e., including his socks).
6) Berger admitted to taking the documents when confronted because he had already reaped the marginal benefits of taking them out of the National Archives. The marginal benefits exceeded the marginal costs when he was inside the Archives and wanted the documents out, but the situation changed after he got them out! In fact, pleading guilty but careless may have just been another ploy to further reduce those marginal costs.
7) Berger removed some documents a second time after being caught. This can only indicate that the marginal benefit of the documents went up again.
8) Berger did not keep the documents because this is when the marginal costs of his actions go through the roof. Suspicion of theft is one thing, but possession of the stolen goods is quite another.

The gist of the argument that Berger would not have taken the documents intentionally relies on the claim that the marginal costs in steps 4 thorugh 6 are so large that they swamp the total benefits at all the other stages combined.

But, this isn't quite correct! Actions 1 through 4 took place in sequence, and all must have had marginal benefits that exceeded the marginal costs at that juncture. But, when we get to point 5, those decisions are sunk, and irrelevant to the optimal choice at point 5. Again, taking the documents and hiding them indicates that the marginal costs at point 5 were not that large compared to the marginal benefits, and the same argument can be made about actions 6 and 7. He stopped at action 8, which is where the marginal costs exceed the marginal benefits, and is also the first point in the sequence of events where total costs incurred might exceed the total benefits obtained.

I hope I've convinced you that the position that this was a careless mistake is untenable. If so, what story can we put together to explain these actions? The one that makes sense is that Berger went to the archives to find something that he knew or suspected was there, and removed it to show to someone who either did not have his security clearance, or did not want to be known as a reader of these documents. Whatever was in them was so important that this happened more than once.

A naive viewpoint would be that since Berger is a Democrat, and until recently involved in the Kerry campaign, that the documents must have been something that would harm the Bush campaign. But, this doesn't jibe with the hiding of these actions. If Berger had a document that would harm the Bush campaign it would make sense to let everyone know about it. But ... that hasn't happened. This tends to point to the flip side of this political equation: to win you either have to score points on your opponents or else prevent them from being scored on your own team. A more sophisticated viewpoint is that Berger was trying to inform Democrat people about documents that were in the Archives that could be used against them.

Here again, the tale gets fishier. It's hard to fathom anything the Clinton administration did with respect to terrorism reflecting poorly on the Kerry campaign or current Democrats. No matter how adversarial you are towards the Clinton administration, it is hard to fathom how they could have envisioned how the past 4 years have turned out. There is no fault in that. So, what do I conclude from this? That the Bush administration closely followed a gameplan laid out by the Clinton administration regarding terrorism, Iraq, Al-Qaeda, and so on. Opposition to those policies is a key to the support that Kerry has been able to garner, and it would be tough for him to continue to sell the position that his followers view as superior if it turns out that he is arguing against both a Bush and a Clinton policy. One other possibility is that it appears that this theft occurred in the fall, so it is possible that Al Gore was the target of the classified information, and this was a way to get him to back out. Pick your poison ....

Comments

[ken anthony]

Does the analysis change if he was actually adding docs to the archives?

http://pittsburghlive.com/x/tribune-review/opinion/archive/s_204906.html

[Dave Tufte]

I don't think that it does.

The point is that the costs outweigh the benefits only in total (rather than on the margin), and then only after he got caught.

Presumably, if he wanted to add something to the records in the Archives he would do so if the marginal benefits exceeded the marginal costs.

This does change my analysis, in that, if he was adding something to the Archives by removing something else, it probably was a modification of what he removed. That's pretty easy to check, as Peter Bacanovic found out the hard way.