Data mining is based upon algorithms written by analysts. It is governed by the same potential for error that pervades every human endeavor. Analytic errors, however, are the least of the problems with data mining. In a typical algorithm, there are many variables that can be manipulated to attain pre-specified objectives. For example, suppose a large property leasing company wanted to increase its profit margin by raising its income. Among the many variables it might consider—cutting costs, reducing vacancy rates with longer term leases, increasing income by building new units in high costs housing regions, automating lease management procedures, etc.—it might include the universal economic element of demand. One of the byproducts of bloating the demand calculation is that it can be self-fulfilling. Market analysis of demand is behind the co- location of so many fast food vendors within a stone’s throw of each other. Interestingly, co-located fast food chains in malls do seem to increase demand because of their ready availability to shoppers. But in the case of our property leasing company, as much as a 20 to 30 percent increase in rents can be justified by simply tweaking the demand variable. Alleged market forces instead of a desire to increase corporate profits can then be credited for gouging tenants.

Now apply the same principles of analytic data mining to politics, specifically to politically based polling. Electorate data collection and interpretation can be effective in directing a political campaign to win its identified target audience. But do you see the potential problem here? The same process can be easily manipulated to influence public opinion—which explains the contradictions between polls conducted by the opposing political parties in the same political contest. In the last presidential election, for example, both parties told the voting public that their candidate polled better and would win. (One Party went so far as to actually believe their own skewed polls.) Having stated this problem, I’m not claiming that all pollsters are fudging the results. I’m sure there are reputable organizations doing their best to cull data they think helpful to politicians, as well as the electorate. But how should we the public evaluate the usefulness of this information? If a majority of polls agree with me, should I feel affirmed in my position? If they disagree with me, should I merely disavow the poll results as biased? Or should I consider the pros and cons of every campaign issue or candidate on the merits of respective policy positions? In other words, should I just ignore the polls?

My last question seems to imply the obvious answer. But the problem I’m identifying cannot be so easily ignored, for it is both pervasive and even subversive. To be succinct, we are living in an era of massive data manipulation where basic trust in institutions, industries and government is being undermined, often by these very players. The goal is not just to inform or even to influence the public. Within the realm of politics, polls can be used to obfuscate facts and positions, making the development of practical policies nearly impossible. Here are some examples:
➣ According to the polls, a majority of Americans agree that reducing deficit spending and its antithesis, taxes, are both necessary.
➣ According to the polls, Americans want to reform the safety net while neither cutting benefits nor increasing the taxes that support it.
➣ According to the polls, Americans support the Second Amendment and gun control legislation. (These positions are actually not contradictory, though they are presented as such.)
➣ According to the polls, Americans believe in global warming, but not in legislative restrictions on the use of hydrocarbons (e.g., the carbon tax) or in major government investment in alternative sources of energy. (Our government’s spending on these alternatives lags behind many European nations and China, even though Americans seem more than willing to install solar panels and drive hybrid cars.)
➣ According to the polls, Americans believe in sustainable development but not in the role of government to shape it. (It seems likely that sustainable development might not be understood in any relatable sense. It was defined a quarter of a century ago by the World Commission on Environment and Development, as meeting “the needs of the present without compromising the ability of future generations to meet their own needs.”)

This list can be extended, but it suffices to make a point. Pollsters can ask questions and manipulate variables to justify positions that are never specified. In other words, the questions answer themselves without ever dealing with actual policy. For example, how would Americans respond to actual legislative bills that would reform specific elements of the tax code, Medicare, and Social Security? Likewise, would Americans support a carbon tax, environmentally sensitive restrictions on development, and investments in alternative energy sources if the costs and benefits to our posterity were fairly presented? Rarely are we debating actual policy formulae. Instead, we seem to lose focus amidst alleged value conflicts. For example, support for the Second Amendment is used as the argument against background checks. The polls that support the right to bear arms are quoted to advance this argument. But these polls are irrelevant to the real problems we face. Recently it has been reported that 30% of the weapons used in the Northern Triangle of Honduras, Guatemala, and El Salvador were purchased at American retail outlets. Do the polls address this issue and its impact on young immigrants from that region trying to escape the violence? And what right of gun ownership belongs to mass executioners here or abroad? Why do we seem unable to form a gun control policy that does not in any way conflict with the Second Amendment? Perhaps we are responding to a pollster’s question that might read as follows: “Do you support background checks that infringe on your right to own a gun and violate your Second Amendment rights?” I have received poll questions with this type of construction. They elicit a desired response based upon strongly held values, but offer no actual policy formula to address the real problem.

I am not a legislator so I probably should not specify possible bills without the debate and compromise required by our system of government. But I think you can see my point: poll data can affirm public opinion as justification for a lack of actual policy. Both sides of the political spectrum can claim public support for doing nothing! Therefore, no legislative action is undertaken. Remember the so-called “grand bargain” that the President and House Speaker had nearly reached several years ago. Its failure seems to be the demarcation between any possible policy compromise and the current situation in Washington. Rhetorical flourish has replaced policy debates. Accusation substitutes for self-examination and accountability. One-upmanship parades as political virtue; and compromise is a political vice. In this surreal context, the polls are used to serve political gridlock and become nothing more than arrows in the political quiver. They justify the lack of policy proposals and the opportunity for any debate on the matter. Without these proposals and relevant debate, there is no opportunity for compromise. Without compromise, there is no policy.

We are polling for non-action. It’s like fishing for dead fish.