“Yes Virginia, There are Alternative Facts!”

20170122_todd1Trump spokesperson Kellyanne Conway was universally condemned for using the term “alternative facts” while defending President Trump’s claims of attendance at his inauguration. Everyone scoffed. Surely there are just facts, falsehoods and unknowns. Indeed when this occurred in an interview by NBC’s Chuck Todd, his face became red and he expostulated, “Look, alternative facts are not facts. They’re falsehoods!

It’s too bad but he is wrong, as are all those who are ridiculing poor Kellyanne.

What is a fact? It is something believed to be true. But that belief can be based on a proof from fundamental principles or it can derive from competent data analysis. Generally speaking, mathematical facts are of the first type, while economic facts are of the second. The certainty of our belief is usually higher for the first type. Interestingly, religious facts are of that type, but they are distinguished by applying non-scientific fundamental principles.

To demean or even reject the second type of facts is a crippling intellectual mistake for human endeavor. Perhaps they could be labeled theories, conjectures or hypotheses, but that would reduce their force and utility both in science and in the public discourse. One would need constantly to apply confidence levels that would be confusing to all but experts, and the resulting convoluted terminology would impede the general understanding.

The second type of facts are conditional, because new data or refined analysis may alter conclusions. Moreover, at any given time there are often reasonable alternative analyses of the available data, leading to – you guessed it – alternative facts! Does this mean that these facts are nonsense? I don’t think so, and neither should you.

The preceding may seem too abstract, so let’s take an example or two. Consider the U.S. unemployment rate. An official announcement by the Labor Department put it recently at about 4.8%. But closer examination reveals that it was computed by ignoring people who have simply given up looking for work, while it includes many who work part-time because they can’t find full employment. It is entirely reasonable to reverse both of those decisions and arrive at a much higher unemployment rate, well north of 10%. These are alternative facts. Neither is a falsehood. They simply represent different analyses of the same basic data.

Even the representation by Kellyanne Conway deserves some reconsideration. Let us suppose that she was including people watching the inauguration on TV. Surely this isn’t an absurd analysis. In this age of technology, actually being present at events is hardly the usual circumstance. Is the audience of the Golden Globes merely those who had tickets to the event? I have read estimates of the watching audience that would actually support Trump’s claim that his inauguration was seen live by more people than any previous one. This would be a justifiable alternative fact. Of course, we don’t know that this was what she meant and it would have been smart of her to say so if it was.

Does this mean that all alternative facts are equally persuasive and deserve equal weight? Of course it doesn’t. One must examine the quality of the data that are used and the applicability and certainty of the analysis. This requires expert review and isn’t something casually performed. In the end, this devolves into a dependence upon expert opinion. That brings into question the quality of this opinion, which is largely a peer-reviewed process. Finding some expert or even small group of experts who believe an alternative analysis is insufficient. Strong consensus is the gold standard.