Neil Degrasse Tyson is a scientist and host of PBS's Nova Science Now. I have enjoyed his program immensely. However, I was quite disappointed in his
op-ed in the New York Times today in which he intimates that Democratic voters are choosing the wrong candidate in Barack Obama. His reason: Statistical analysis and computer modeling of polling data show that if the election were held today Hillary Clinton would win the general election against John McCain and Barack Obama would lose.
The computer model of polling data Tyson refers to was created by two astrophysicists, using data they gathered prior to the 2004 general election. What Tyson, also an astrophysicist, does is apply this model to the primary election. He uses a complicated system which involves the use of polls in states where they are available, and the use of the 2004 election results when there are no polls available.
His premise, analysis, conclusion and question are so absurd that I have to refute each one of them, not from the perspective of an astrophysicist, which I am not, or from the perspective of a political scientist, which none of us are, but from the perspective of an expert in human behavior, which they are not and I am.
Tyson says: "One of the tasks of scientists is to clarify the apparent complexity of the universe by using the language of mathematics." This may be true in the hard sciences, but not the soft sciences, like psychology and political science, where the unpredictability of events and human emotion have profound effects. While some human behavior is predictable, much of it is not, and in politics, as we have observed, unexpected things happen that change a voter's opinion over the months preceding a general election.
Tyson's premise is that we can use statistical analysis of past events to predict future events, not of how the stars and planets behave, but of how humans behave. Since human behavior is always changing, and events humans must respond to are always in flux, this is a false premise.
Tyson's analysis is flawed in that he analyzes data preceding a primary election with a computer model that has been developed for and used prior to a general election. Not only is there a time difference here, in that many events can intervene to affect voters' minds between the primary and the general election, but there is also a different level of attention. Many more people are paying attention to the candidates prior to the general election than they are prior to the primary elections. Furthermore, it is absurd to think that a model used to explain what happened prior to a general election, based on polls preceding that election, can be used to explain what will happen in a future general election, based on polls taken before a primary election. Dr. Tyson is not only comparing apples and oranges, he is taking information about this year's apples and applying it to next year's oranges.
Tyson used "the simple rules established by Professor Gott and Dr. Colley: in states in which a poll has not been taken, you give that state to the party that won it in 2004. You do the same for states where the median poll is a tie." Here there is a flawed assumption, an assumption that nothing much changes in states between elections. Red states stay red and blue states stay blue. Political observers are already seeing a change in the electorate as they characterize more and more states as "purple," meaning more states are up for grabs than ever before. You cannot use the 2000 and 2004 electoral maps and expect that it will be repeated with any certainly, especially considering the work Howard Dean has done in the interim, implementing a 50 state strategy, and the work Obama has done in making inroads into formerly red states.
Finally, with respect to Tyson's analysis, there are factors in this race that have never existed before. We have never seen a president with a 25% approval rating, nor 81% of voters who believe the country is on the wrong track. Nor have we ever had a major party nominee other than a white male. Surely, these factors could confound the results of any computer model.
Since Tyson's premise and analysis are flawed, it stands to reason that his conclusion is flawed. Hillary Clinton would win the election if it were held today, he says. I don't think his mathematical model comes close to proving that, but even if it did, so what? It says absolutely nothing about who will win five months from now, when the public has gotten to know the candidates better. And herein lies another problem with Tyson's analysis. The voters are far more familiar with Clinton and McCain than they are with Obama. That will be different in five months.
What will also be different in five months is how the Republican smear machine will or will not have hurt the Democratic candidate, or will have backfired. Tyson may think the smear machine against Obama would only make a Clinton victory more likely, but he is not considering the smear machine the Republicans would gin up against Clinton. Can he possibly forget the ugliness of the impeachment proceedings against her husband? Can he honestly think the Republicans would not use this against Clinton? Is he unaware of Bill's secret business dealings and undisclosed contributions to his presidential library? The Clintons have baggage, and there is no doubt in my mind that McCain's team would unload on her, lessening her chances for victory.
And while this may seem out of the question to Tyson, it is possible that the more people get to know about McCain's voting record and policy positions, the less they will like him. This long Democratic campaign season has kept the focus off of McCain. Now that we have a Democratic nominee, it is likely the press will finally give some attention to McCain, and the multiple mistakes he has made that have stayed under the radar will start coming to light. McCain's higher poll numbers against Obama are in part a result of his not having much press scrutiny to date. That, of course, will change.
Now to Tyson's question. He asks "what does it say of the Democratic delegate selection system when its winner would lose the presidency if an election were held today, yet its loser would win it?" I can hardly express how ridiculous I believe this question to be, and how inappropriate it is to ask it after the nominee has already been chosen. I have to say it makes the good professor sound like a die-hard Clinton supporter who is disgruntled that the young Obama defeated her. Nevertheless, I am going to attempt to answer his question.
"What does it say"?" It says that Democratic voters cast their ballots on the basis of something other than mathematical models. It says that Democratic voters look at the candidates and vote for the one they want to see in the White House. It says that Democratic voters this year decided they wanted the change candidate more than the candidate who already spent eight years in the White House, married to a man who shamed the Party and the nation. It says that Democratic voters have hope, and believe in the ability of a gifted young politician to win over voters before the general election, just as he had won over voters in all the states in which he vigorously campaigned.
Four years ago, a majority of Democratic voters decided to vote with their heads rather than their hearts. While there was tremendous enthusiasm for both John Edwards and Howard Dean in 2003 and early 2004, by the first caucus in Iowa voters had decided to go with the candidate that they thought could defeat Bush, even though he didn't excite them. They may not have had a statistical model to guide them, but they were sure Kerry's Vietnam War hero bona fides would ensure his victory over Bush, who had been AWOL during part of his service in the Texas Air National Guard, which daddy Bush got him into so he wouldn't have to go to Vietnam. They picked the man they thought was a sure winner, just as Tyson is sure Clinton would be a sure winner. And as we all know, Kerry lost.
This year, the Democrats had two strong candidates, a woman and an African American. Both are the first candidates of their gender and race to get this far, and the candidacy of either one of them, because of that, would confound pollsters and statistical models.
This is a different year. People are looking at an entirely different kind of race than the fifty-four previous ones. We have no idea what might happen. But some of us, those who voted for Obama, are convinced he can overcome the misinformation about him, the prejudice against him, the old red-state, blue-state divide, and the conventional wisdom which guides pundits and even, apparently, astrophysicists.
Would Dr. Tyson really have us vote as robots, using computer models to determine our vote, or would he rather allow this to proceed as a real contest for the hearts and minds of the American people, a process that has a great deal of suspense and uncertainty associated with it? If Democracy means anything at all, it means that we choose our candidates on a number of factors, that we expect our candidates to have a conversation with us to help us choose, and that we are able to make up our own minds. We have five months ahead of us for this process to unfold.
If Dr. Tyson cares at all about this country, he might start by encouraging people to pay attention to the issues as well as the skills of the presidential candidates and not lament that the Democrats did not use a statistical model to choose their nominee. And were he to study disciplines other than astrophysics, he might learn that on this planet, and in human behavior, nothing stays the same. There are unpredictable forces that always upset mathematical models, and one of them is the human will.