Butterly Flappings

Edward Lorenz is something of an accidental historical figure, thrust from the pedestrian corridors of meteorology into the sexy and mysterious world of chaos theory.  The guy took a short cut because he wanted to get a cup of coffee and ended up sort of famous.

It was 1961 and Lorenz was running some computer weather models.  As you can imagine, processing speeds were a little slower then since vacuum tubes were involved.  He interrupted a scenario to change a particular input value but entered 0.506 instead of the more precise 0.506127. No big deal, right?  When he got back from break, the computer had simulated two months of weather, but the results were wildly inconsistent with what he had expected.  All because of that minor change in one variable, and thus was born the idea known as the ‘butterfly effect,’ the notion that a butterfly could flap its wings in Brazil and cause a tornado in Texas.

There is a lot to this chaos theory (the scientific one…life with toddlers and teenagers is ‘all chaos, no theory’), but one of the key ideas is that ‘minor perturbations’ in the inputs can result in dramatic differences in the outputs.

This is important to keep in mind as people – proponents and adversaries – begin to crawl into the math behind Uncle Bernie’s ‘Medicare for All’ plan.  As Mr. Sanders, the country’s favorite curmudgeon progressive, continues to sit high atop the early polls for the Democrats, his signature idea is starting to get more attention.  Finally, some people are starting to ask, ‘Well Bern, how much will that cost actually?’

OK, the short answer is a mind-numbingly large number.

But how large depends on where you set some of those little input variables.  Estimates, and at best these are back of the napkin, range from 12.5% to 21% of GDP. 

Four variables, none of which are actually small, are:

  • How much do you pay providers?
  • How much savings do you get in prescription drugs?
  • How much administrative cost is required to run the system? 
  • How much will utilization go up if everyone has coverage?

Let’s take a minute on the first one, the one we all here care about most.  Pay docs and hospitals the same as current Medicare rates and you’ll send many hospitals, particularly in rural areas, straight into insolvency.  Pay more to reflect real costs and the price tag soars on what is an already difficult political sale.

So, as we begin to contemplate the biggest proposed policy change in generations and the numbers start flying around, keep the lesson from Mr. Lorenz handy.  Minor perturbations to big complex models have wildly unanticipated impacts.