Quest for Power: Politics in America 2022

One of the greatest evenings of my life was Saturday, November 11, 1972. I was honored to be at a dinner party of eight hosted by John Kenneth Galbraith, for whom I was then the head teaching assistant at Harvard. The guests of honor were Senator and Mrs. George McGovern, who that Tuesday had lost the Presidency to Richard Nixon in a landslide, carrying only Massachusetts, which fittingly was where we were. I was an unabashed liberal and was honored to be in his presence. I have to confess that only eight years earlier I had supported Barry Goldwater and gotten Richard Nixon’s autograph at a campaign rally (he signed it “Dick Nixon” which he never did once he became President).

Then I entered the real world and began to work in Washington, something I did off and on for the next forty years. The ends I sought and supported never changed, really: they are still largely the same as I had fifty years ago. What I learned was that not all means work. There are many things government could do with the right staffing and the right incentives, but we as a people guarantee that we will have neither.

My first in depth lesson came over aircraft carriers. I was on a team building a complex, nonlinear simulation model for naval combat. It was, especially for the time, an incredible intellectual achievement. Feed in the parameters of the war you needed to fight, and the model gave you the optimal force structure to fight that war. In particular, the model forecasted how many carrier groups the Navy required. Everyone on the team was proud of the work we had done.

The client said it was great, but they asked for a complete and exact copy of the model, including the computer on which it ran as well as all the software and data. We said there was no need, and they said they were the customer. We delivered it to them. We were not exactly sure why, but they were the customer.

Each year as part of the budget process, each military service submitted their estimate of the wars they were prepared to fight as support for their budget request. This estimate was the input to our model. It was mind-numbing detail. I have been building simulation models for decades, and I am not sure why but most real world models end up being incredibly sensitive to some seemingly minor inputs.

Popular media called this the butterfly effect: the idea that a butterfly flapping its wings in China would eventually affect the weather in New York. A mathematician…