I was reading something by John  Fund¹ this morning when I realized that political insiders sometimes lose some degree of common sense. To be specific, the pundits sometimes forget that "the voters" is not some entity that exists out there, responding in mechanistic ways to various inputs, but is a huge mass made up of thinking individuals just like them.

You hear it all the time "The Democrats have more money, so the Republicans are in trouble", or "the Democrats ran right, so they outflanked the Republicans". All of these statements contain a grain of truth, but all take that little truth and take it too far, forgetting that elections are not mechanistic, and that money or policy statements alone do not win elections.

Take, for example, money. A campaign can obviously be lost for want of money, but that only goes so far, at some point money is adequate and the value of additional inputs is small or nonexistent. Yes, it is important to get our a message, but at some point the value of money stops. In fact, sometimes having too much money can hurt a campaign, just ask Mitt Romney. Being the guy with all the cash does not always play well with the voters, nor do voters like to be inundated with too much advertising. If they feel they are being manipulated, voters may rebel against the advertiser.² But sometimes pundits miss this by concentrating on elections as finite state machines, determined by a number of "factors".

The same can be said for "triangulation". While it is true that Bill Clinton was elected and reelected by "running right", and it is also true that some Democrats in 2006 won by shifting right, they could only do so because they managed to appear to sincerely believe in their "centrist" position.³ Voters are not idiots who will buy whatever a candidate says. When Hillary plays "hawk" on Iraq, does anyone truly buy it? Voters are not brainless, and do not lack memory, they know Hillary is trying to look centrist, but has no love of the military and will cut and run the first moment she can do so. Her efforts to appear hawkish may convince voters she will be less likely to retreat immediately than her rivals, but that is all.

The same applies to the recent Democrat stand against earmarks. While it may play well to the press, the electorate does not buy it. They know that, of the two parties, the Democrats are the more besotted with taxing citizens, and any sort of spending restrictions, however unlikely it is either party will support them, will come from the Republicans. Just because the Democrats mouth fiscal responsibility does not mean the voters will buy it. To convince the voters, a candidate needs both a compelling position, and credibility. Pundits often forget that second part.

In some ways, I blame our social sciences. All our social sciences seem dedicated to abstracting the humanity from human behavior. For example, it is rare to read an economics text which speaks of "choice", all postulate curves and charts without explaining where they originate. They seem to imagine "the market" as some vast deterministic mechanism, without any human involvement. Only a very few economists have even stated the obvious that, due to human nature, those "supply and demand curves" are not fixed structures, but can fluctuate wildly from second to second.⁴ Most traditional economists treat them instead as rather static "givens", with no reference to the human choices underlying them.

And economics is hardly the worst offender, one need only read a sociology text to see the strange spectacle of a discipline dedicated to human behavior which seems to completely ignore individual choices. Social science in general, for whatever reasons, has somehow forgotten that the rest of humanity, just like the social scientists, make decisions for themselves, based upon a host of unknowable influences, not as the result of a fixed, mechanistic process driving "society" without reference to any individuals.⁵

But we remove the humanity from human behavior at our own risk. When we try to predict elections based on "factors" driving a deterministic model, without giving any thought to the humans who make the final decision,⁶ we often arrive at proven, sound, rational conclusions which turn out to be absolutely incorrect.

But anyone who has followed the predictions of political pundits know that to be true.

-------------------------------------------

1. To be fair, Fund is actually one of the better commentators and tends to analyze elections relatively thoroughly. It was simply his mention of Republican money troubles in congress which brought this whole topic to mind, so he has the misfortune of having his name attached.

2. There is also a limit to what money can do. Some campaigns are doomed regardless of money. No matter how much you pour into a "Duke for President" or "Farrakhan for President" campaign, you have no hope of winning.

3. The same applies to Bush's "compassionate conservative" position, which essentially amounts to a Republican running left. It wasn't mentioned at the time of the election, but Bush basically won by triangulation as well.

4. VonMises' Human Action is one of the rare exceptions, as he spends a large part of the text on explaining how human decisions work, how all valuation is relative to other values, ordinal not cardinal, and then showing how those decision making processes translate into the phenomena of economics. But he is unusual in this regard.

5. Actually, the "all humans are machines" assumption is what always puzzled me about behaviorists. Do the behaviorists themselves think they are machines as well? Or do they think they are the only truly "rational" beings, and the rest of us are machines? I could never quite grasp the thinking of the truly consistent behaviorist.

6. Yes, we can still generalize about humanity. There is still some regularity to behavior, with a bit of random deviation. My complaint is not that we should never generalize, but that sometimes our models serve to obscure the truly human factors. For example, a model may predict that, given enough money, Obama could win over the klansman vote, where a bit of rational thought would show that to be impossible. While such a stupid result would rarely make it past even the most obtuse pundit, only slightly less stupid conclusions do make it through on the strength of "models" and :theories" every day.