Working Up

by Dwayne Phillips

Estimating budgets months and years into the future is difficult. I have yet to meet anyone who does it well enough to risk large amounts of anything valuable. The Excel spreadsheet calculates well enough, but it can provide a false sense of precision. Instead, I use a slide rule. The difficulty of using it and the precision it provides keeps me humble.

Some people in my office were estimating budgets this week. They were doing their best to predict the future. This isn’t easy, but I believe it is necessary to attempt. Planning (predicting the future) helps us once we arrive in the future.

When people work with numbers in today’s business, we use a spreadsheet. Excel, from Microsoft, is what the vast majority of people use. Excel works as well as any other spreadsheet. A calculator would also work, but then you have to type the numbers several times to put them in a computer report and all that punching of numbers is error-prone.

What I don’t like about Excel and budget estimation is the false sense of precision we have. We can type in some numbers, enter a formula, and there we have it – we need 33.333333 man hours of engineer time and $4,327.188762 on May 13 of 2011.

Now I am overcome by the silliness. All these decimal points. All this precision (I emphasize “precision” and not “accuracy.”) Do we really know what we are doing? Are we betting the future employment of people on the 9th significant digit?

When estimating budgets, I use a slide rule. I bought a pocket model about ten years ago and I keep it on my desk. I emphasize the pocket model as it is small and I strain to see the little lines even with reading glasses. This reduces my precision in calculations. If you aren’t familiar with a slide rule, see the Wikipedia page for an introduction. A slide rule is a calculating device that provides about three decimal places in every calculation. So, for the above example, I would estimate that we need about 33 man hours of engineer time and $4,300 in mid-May of 2011. If I were being candid, I would further reduce the precision to 30 man hours and $4K in the Spring of 2011.

Predicting the future isn’t easy. Convincing myself that I can predict the future with great precision won’t help me. The false precision will probably bolster my confidence beyond all reasonable measure. I don’t think that will help me either. I suppose it is possible to program Excel to throw out the extra decimal places and show answers that are as precise as is reasonable. I, however, prefer to use the slide rule. There is something about it that brings humility to the calculating task.