[ This was originally posted as "Will Candidates' DNA Play a Role in Future Elections?" at AlterNet. ]
Moments after Barack Obama was elected, pundits began to speculate about future elections. And for once there is a new idea to add to the horse-race discussions: some commentators have proposed that we look at candidates' genomes to discover if they have the qualities we need in a president.
Personalized genome scans will almost certainly be better, cheaper and much more widely available during the next presidential election cycle. It therefore makes sense to discuss this possibility now, outside of the partisan frenzy stoked by campaigns. The Wall Street Journal, the New England Journal of Medicine, and the New York Times have all given space to the idea, and the authors have taken very different positions.
Would an inspection of candidates' DNA be helpful? Could we learn enough about their health, character and ability to justify this intrusion into their genomic privacy? Should anything be off-limits? These are valid and interesting questions with important policy implications. Considering them in the context of a hypothetical election can also teach us something useful about how and when to apply the results of genetic tests to our own everyday lives.
Some enthusiastic researchers, notably the Personal Genome Project's George Church, think we should scan the DNA of all candidates, and publish the results. Public health scholars Robert Green and George Annas, on the other hand, warn against the possibility of "genetic McCarthyism." They're concerned that DNA results would be abused as a new form of opposition research, with dire and misleading warnings being broadcast in attack ads: "Can we risk as President someone who may [perhaps, eventually] suffer from a [potentially] debilitating disease?"
Even worse is the prospect of someone being asked to "prove" their racial purity with a genetic test -- and this has already happened, in Turkey. President Abdullah Gul is considered by some in the far right to be "soft" on Armenians because he has refused to condemn calls for an apology for the 1915 ethnic cleansing, which is widely regarded as genocide. One politician has called for him to demonstrate that he is not part-Armenian: "These days, scientists use DNA tests, not family trees, to identify ethnic identity."
Such an overtly racist abuse of testing may seem far-fetched here, though there are already tests that purport to demonstrate membership in particular Native American tribes, and indeed to show Jewish ancestry. More pressing is the possibility of misleading medical prognoses, and an early defense against this prospect may be better public understanding about what genomic tests can and cannot do.
In a sense, we have always used genes, very crudely, to help us choose our elected leaders. Two pairs of fathers and sons have held the highest office, for instance, and three Kennedy brothers have run for it. There are many other examples of families with several members elected to Congressional and other offices; the Udalls include two incoming Senators as well as another cousin who just lost his seat and several distinguished ancestors. Burke's Peerage is said to have claimed that every presidential election "has been won by the candidate with the most royal genes." Certainly, there are other factors -- policy can make a difference -- but some people do look on genetic inheritance as a qualification.
However, genomics is a science of statistical possibilities. Genetic tests can almost never tell us with certainty whether someone will come down with a debilitating disease in the next four or eight years. A few medical conditions can be predicted with real confidence from genetic data, but in almost all cases the most that can be said is that there is an increased or decreased chance of hypertension or cancer or some other disease.
Analyzing the effects of tests for multiple conditions complicates everything much further. If you have a 50% increased chance of heart disease but 40% less than normal likelihood of Alzheimer's and 30% less of Parkinson's (these are made-up numbers, as a simplistic example), is that on balance good or bad? Arguably, it's quite hopeful, since you can take effective preventive measures against heart disease through diet and exercise. But what if you have to consider twenty or fifty genetic predispositions: how do you factor them all in?
This demonstrates a major problem with generalized genomic tests, especially if they are to be used for such a complex question as choosing an individual for a particular job. George Church is missing the point when he says that "it is not like we are collecting horoscope data or tea-leaf data. These are real facts, just as real as bank accounts and the influence of political action committees or family members." What do the facts mean? How do you interpret them? The date of birth you give an astrologer is factual; the predictions are a matter of interpretation.
What any particular collection of genetic variants implies is a very tricky question. It's so difficult that the Departments of Health in California and New York have complained about direct-to-consumer genetic tests precisely because consumers do not have the expertise needed to evaluate the data without expert help. They have argued that companies selling gene tests are, in effect, practicing medicine without a license. (The best-known California firms involved have settled this dispute, but others have limited their activities.)
So how would voters and commentators interpret a candidate's gene scan? More than that, how would we make political judgments even if the genomic facts were more or less clear? President Lincoln may have had Marfan Syndrome, which could perhaps have been predicted from a genome scan. He may also have suffered from a form of depression that might have been reflected in his DNA. Would we have been better off disqualifying him? Would the nation even exist if we had?
Certain medical conditions are generally considered important for voters to know -- but even then, the case of Franklin Roosevelt muddies the matter. Hard as it is to imagine nowadays, he concealed from the public the fact that he was paralyzed and used a wheelchair. And he was elected four times and is generally ranked as one of the three best Presidents ever.
What was FDR's greatest strength? Many say it was his temperament. That may be partly inherited -- there are plenty of phlegmatic children of phlegmatic parents -- but there is no known specific gene for temperament, and therefore no genomic test. Even if we could test for a collection of genes associated with calmness in the face of crisis, the answer is very unlikely to be definitive. Stanford's James Gross has pointed out that "genetically identical people can give very different outward impressions because they think differently, they regulate their emotions differently." Genes, in other words, are only part of the story.
And sometimes what the genomic data suggest is just flat wrong. For instance, one anemic woman was surprised to discover that she had a gene for hemochromatosis, which involves abnormally high levels of iron in the blood. (Sounds like a country song: Who you going to believe -- the DNA print-out or your lying blood?) James Watson, the co-discoverer of the double helix and one of the few whose genome is public, does not suffer from either of two diseases for which he "has the genes."
Balancing the possibilities revealed by genomics is never going to be the best way to select a President. And learning how to interpret such tests for ourselves, for our own use, will itself be challenging. Sequencing technology will continue to improve; predictive interpretation will too -- but if you're looking for certainty, don't hold your breath.
Meanwhile Secret Service agents bag and trash any glass the President uses while away from the White House, so no one can steal his genetic secrets. It's probably just as well.