I worked for Ira Magaziner during his health care reform campaign. I came out of there thinking that there is no such thing is "too stupid to be true": things are always less rational and less organized than you would believe...
With respect to Russia... The U.S. Treasury Department's Office of the Assistant Secretary for International Affairs has about 300 people working for it. Half of those work on issues relating to the other industrial economies. The others are--if I recall correctly--divided more-or-less evenly between people watching Japan, people watching the rest of East Asia, people watching the Middle East (oil, you know), people watching Latin America, people watching Africa and South Asia, and people watching Eastern Europe and the former Soviet Union. Of those watching Eastern Europe and the former Soviet Union, one-third or so deal with stuff going on in the former Soviet Union.
That's eight people. Probably two secretaries, three economists, two lawyers, and a boss. Some of those eight are probably very smart and very knowledgeable. Some aren't. There may be useful experts at the State Department as well (but I doubt it: State is still filled with experts on the nomenklatura and on mutual assured destruction). There may be someone who knows something about the former Soviet economy working for the Council of Economic Advisors or the National Security Council (but maybe not). There are a lot of people who know useful stuff at USAID, but they are "line" not "staff"--which means that they very quickly fall into the pattern of defending whatever programs are ongoing, rather than being willing to sit up and say "this isn't working, we need to do something else."
The upshot is that the "big research department" is smaller than you would believe. The only thing that could potentially save the situation is that those eight can draw on practically anyone else in the world they want to talk to. But that leads to good outcomes only if the eight know which of the many people seeking their ears they should trust, and which questions to ask...
Brad DeLong