The Pakistani Spectator

We all have in our heads ideas for how the world works, ideas about human nature, about politics and economics, ideas that we believe explain the world, or some part of the world, more or less. There is some satisfaction in feeling we understand the way the world works, that we know “truth.” Most models of how the world works, however, are conditional – they don’t just say what will happen next – they predict likely consequences that depend on the actions people choose. These models, if they are at least roughly true, can help us choose our own actions to achieve future results that we prefer to alternative possibilities. These models can be very elaborate and detailed, like Adam Smith’s model of the free market, or they can be simple rules of thumb, rules like “Arranged marriages work out better than marriages between young adults who meet on their own.” (I suspect that this latter example is true in some cultures, and not true in others, but I mention it just as an example of a fairly simple model.)

I know that first paragraph sounds terribly vague, so I’ll give specific examples:

A capitalist model of the economic world predicts that a relatively unregulated market will produce the greatest economic good for the greatest number. It also predicts that someone who works out how to deliver a product for a lower cost than any competitor can prosper in a capitalist market. An individual who believes in the capitalist model of the economy would choose specific actions that would tend to bring him or her more a more favorable future under the assumption that the capitalist model is true. For example, he or she would vote for government leaders who will avoid excessive market regulation and unnecessary government ownership of economic resources. Furthermore, he or she would personally look for ways to maximize personal economic productivity, knowing that the easiest path to riches is to produce as much economic value as possible.

On the contrary, a communist model of the world predicts that economic output is relatively fixed, and the happiness of all is maximized when the fixed economic output of the nation is spread evenly to all, something best achieved by government ownership and control of all important economic resources. An individual believing the communist model of the economic world would work to eliminate a government favoring a free market. An ambitious communist would further tend to focus efforts, assuming selfish desires for advancement, on achieving a high position within the government-run hierarchy. Efforts to enhance personal productivity would be wasted energy, except to a saint, in an economic system that predicts equal economic rewards regardless of one’s efforts to produce more.

Communist and capitalist theories of economics are not just different ideas of what is true; they are different methods of predicting what happens if we make different possible choices, and these different predictions effectively prescribe different choices to work toward future results that we desire. A better model of how the world works is not just a satisfying bit of truth to idly contemplate – it is a map for how to steer our lives toward a future we desire. Having the right map to steer by makes a huge difference in maximizing our chances of success, in my view. Therefore, our choices of models for how the world work make an enormous difference to our future successes, both as individuals, and as nations, and reasoned discussion of these models matters. Today’s article, however, is not about specific models of how the world works (except as examples), but about how to choose between models, so that you may choose the best ones.

In choosing models of the world, models to use to guide us when we choose our actions, I think there are three dimensions of those models to consider:

1. Is the model likely true (and how likely is it true)? Most arguments about political theory focus entirely on this question, but I think the other dimensions are more important that most people appreciate. If we are humble, we generally recognize that there are decent arguments in favor of many alternative ways to model the world, and the best we can do is say that we believe one model to be more likely to be closer to the truth (as known only to a perfect being) than some other alternative. Such a belief might be very strong, if we feel we understand the complex issues very well (or if we are simply overconfident!), or that belief might be weak, where our views of the alternative arguments lead us to estimate nearly equal probabilities that each of two alternatives might be true.
2. How useful is the model? Lots of models of how the world works are potentially true without being particularly useful! For example, a completely fatalistic model of how the world works might be true, but it would predict only that any effort to influence the future is utterly futile, offering no map at all for how to influence the future for the better. Further, models can be only approximately true, with known ways they are false, while still being extremely useful. Engineers, for example, know that the formulas of Newtonian mechanics are only an approximation (a very, very good approximation, fortunately!) for the truer formulas of Einstein’s relativistic mechanics, but engineers daily design machines, buildings, bridges, and dams using the formulas of Newtonian mechanics, which is vastly easier to use than the super-complex formulas of relativistic mechanics, which is utterly impractical for solving most ordinary design problems. A theory can be potentially true while being more or less useless because the theory is fatalistic, because the theory is vague, lacking in necessary, useful specifics about exactly what will happen in various alternatives, or because the theory requires more calculation or more data to predict an answer than we can afford.
3. How sticky is the model? Here I am inventing a bit of jargon, “stickiness,” to describe something that doesn’t have a word, yet, as far as I know. Some sorts of ideas are likely to be popular, within a given culture, whether or not they are true or useful. I’m using this new, invented bit of jargon, “stickiness,” to describe this tendency for an idea to be believed regardless of its truth or usefulness to the believer. (Of course, a sticky idea may be useful to some entity other than the believer, some entity for example who promotes the idea and makes it sticky, for example, a bad government that promotes its own legitimacy in public schools.) There are many ways an idea can be sticky without necessarily being true. For example, it appears that young children are genetically programmed to tend to believe what their parents tell them, probably because such an instinctive tendency promotes the children’s survival on average. Therefore, the sorts of ideas that we learn from our parents tend to be sticky, to be believed whether they happen to be true or not, and whether or not we have carefully thought through the reasons we believe them. It is in parents’ interests for their young children to believe that the parents love them and that the children should obey them. A model of the world where we believe as children that our parents love us, and that we should obey them, is therefore sticky. It is also, in most families, true, I think, an example of an idea that is both sticky and true and useful. In many cultures, it is also widely believed that things work out best when adults, also, generally obey their parents, another idea of a sticky sort of idea that parents are likely, in their own interests, to pass on to us at a young age, but in this case, perhaps, not so obviously true and useful as the idea that young children should obey their parents. Several factors can make ideas sticky – the ideas can favor the interests of those who spread them, for example, parents or governments in charge of education, while not necessarily favoring the interests of those who believe these ideas, or the ideas can be flattering or comforting to those who believe them, again without necessarily being useful or true.

I know this is a long post, so if you prefer to read this in bite-sized pieces, this is a good stopping place, and the following three sections can be read one at a time if you come back to the post. I’d like to discuss these dimensions to consider when evaluating alternative political models in more detail, one at a time:

Truth

The truth of an idea, such as a political idea that predicts how well we will do with some particular set of political choices, would seem to be the simplest attribute to discuss, since political debate already tends to focus on truth. Political ideas, however, are usually impossible to either prove or disprove, making for endless debate and little agreement. I think the best thing to keep in mind is that someone promoting a political theory can tell a story where the story appears to demonstrate the correctness of the political idea. However, stories can seem very believable for reasons that have nothing to do with how likely they are to happen. For example, I can tell a very simple and short story, such as “Over the next forty years, the world economy will grow an average of 3% per year, in real productivity per capita.” A better writer than I can write an entire novel set in the near future, with detailed interactions, political movements, wars, assassinations, et cetera. Which will seem more believable, to the human psyche? Instinctively, we tend to believe stories with rich detail, where the details seem to add up to a coherent, well-made story. When we look at those details, however, from a more scientific perspective, every detail is a place where the story can go wrong, and the chances of the future unfolding in a way that even roughly follows a novel are truly tiny, while my uninteresting one-sentence prediction of economic growth, without sounding persuasive at all, has a decent chance to prove correct, precisely because it lacks the detail that adds to its chances of failure.

As an example, Marxist theory is full of predictions that the initial overthrow of capitalism must be violent and must be followed by a period of rigorous government control necessary to overcome counter-revolutionary tendencies to revert to capitalism, but this temporary period of heavy-handed state control will be followed by an everlasting “workers’ paradise,” when government becomes almost unnecessary and personal freedom is maximized. Somehow, in real communist governments, we never got past the phase of rigorous state control, so this theory was too detailed to prove correct. Capitalist theory, on the other hand, is comparatively lacking in detail about the future. It predicts that under free markets, economic output will tend to grow (at least as long as depletion of natural resources does not create a major economic bottleneck), and competition will tend to drive companies to minimize their costs of production, while always struggling to maintain competitive advantages and higher profits versus other companies that are trying to do the very same thing in their market niches. Capitalism further predicts that shortages will tend to correct themselves, if prices are unregulated, as shortages drive up prices, driving down demand and encouraging higher production of the item in short supply, or of any replacements for that item. It makes no detailed prediction of future wages for specific occupations, of the distribution of the increasing economic output (which might be spread fairly equally, if all occupations were in fairly equal demand, or might be highly concentrated in the hands of fairly few persons who possess most of the capital and high-demand skills). Capitalism further makes no predictions about the evolution of governments. The much more limited predictions of capitalism have proven more accurate, largely because they were more limited in their scope, although for a long time it could have been claimed by the Marxists that the writings and detailed predictions of Karl Marx directly influenced far more people, and more governments, than the more modest writings of Adam Smith on capitalism. (Governments following capitalism were mainly just allowing things to follow the natural course they already followed, rather than “following” Adam Smith, who mainly explained the capitalist system that already existed in much of the world, rather than prescribing anything new.) I would argue that the same rich-seeming detail of Marx’s writing both doomed his predictions to failure and paradoxically made those predictions seem more believable, to many non-analytic readers, than the less-detailed predictions of capitalism.

In general, I think we should be suspicious of the truth of elaborate models that make complex and detailed predictions that are based on a foundation of many unproven assertions. These models tend to seem good, precisely because they are so detailed, but their detail makes them far less likely to be completely true than more modest models that claim less.

Stickiness

Does a model serve the interests of the person or group (a school, funded by a government, for example) that taught you the model or otherwise promoted the model? Does a model flatter you, or your nation, your cultural group, your religion, your race, your gender, or your economic class? Does it excuse or glorify historic actions, or the current comfortable status quo? Does it excuse comfortable inaction, rather than demand frightening change and place on you uncomfortable new moral responsibilities? If you answer “yes” to any of these questions, the model is likely sticky, something easy or tempting to believe regardless of whether it is true. (Answers to these questions will depend on who is asked – to most Americans, “Christianity is best.” is a sticky model, while most Pakistanis would find “Islam is best.” sticky! Answers may also depend on when the question is asked – before George W. Bush got America mired in Iraq, the idea that a war with Iraq would help the US was not particularly sticky, but after a few hundred American soldiers died there, it became paradoxically much harder for Americans to believe that the whole affair was a horrible mistake, in spite of the fact that rationally speaking, every life lost in Iraq should have added to the argument against the war! Compulsive gamblers make a similar mistake – the more money they lose, the harder it is for them to walk away from their losses, to admit that the game was a mistake, and to quit the game. In recent American presidential elections, we have also seen an ugly tendency for candidates to be criticized when they have shown in their political careers that they can change their minds on an issue, as if wise people must never have made a mistake, or, worse still, as if wise people must never admit a mistake, and should instead pretend to believe what they no longer believe, simply to appear consistent! This sort of politics has the unfortunate effect of making all political positions sticky to the politicians who hold them, even when the politicians figure out that they were wrong, because they fear criticism of their “flip-flopping” on the issues.)

Although many sticky models are false, the mere fact that a model is sticky does not mean it is false! There is nothing wrong, per se, with a model being sticky, and many sticky models are undoubtedly true. However, when the model is sticky, there is surely a danger that you believe in the model without having given fair consideration to competing, less-sticky models, and, if finding truth is your goal, those competing, less-sticky models probably deserve more consideration. Given equally good arguments for both a sticky model and a non-sticky model, nearly everyone will “buy” the sticky model, and will assign it a much higher likelihood of being true than it deserves. To see the true likelihood of non-sticky models, compared to sticky models, we probably have to “bend over backwards” to consider the non-sticky models, to try really hard to resist the temptation to favor the easier, more comforting, sticky alternatives. A true model may be not just non-sticky, it may be anti-sticky. For example, in the American South, before the Civil War, the idea that slavery was a great evil that spread enormous harm and created great danger to America and to the South, itself, was entirely true, but also very anti-sticky. To believe this extremely inconvenient truth, Slavery-era Southerners would have had to acknowledge a horrible injustice behind their entire way of life, acknowledging evil done likely by their own families and perhaps by themselves. Their moral responsibility to act to correct these evils would have been a heavy burden, indeed.

In a more recent example, Al Gore won a Nobel Peace Prize promoting the idea of another very anti-sticky model, in many lectures around the world, and in a movie well-titled An Inconvenient Truth, where he argues that rich nations in general, and Americans in particular must begin soon to drastically reduce their burning of fossil fuels (oil, gas, and coal, which cause global warming), else risk vast and possibly irreparable damage to the environment and huge future economic damages, to say nothing of moral responsibility to citizens of poorer nations that will suffer for our consumption. We Americans love our cars, and our big, well-heated, well-air-conditioned homes, which are a major part of most of our daily lives, so the thought that we must rapidly change our very way of life, or at the least be prepared to pay a much higher cost to maintain that way of life in some as-yet-not-fully-understood way that requires vastly less use of fossil fuels, is indeed a very inconvenient, anti-sticky truth! The scientific arguments in favor of this truth are very persuasive, but they must overcome the powerful tendency for citizens of rich nations to believe the much stickier alternative model that everything will work itself out if we continue with current practices, or at the very least, that we ought to do nothing until we study the problem (postponing inconvenient action) for a few more years (then, again, a few more years after that, then a few more years after that,…!).

Usefulness

To be useful, a model should not simply explain the past, and perhaps the present, but it should make plausible predictions about the future. Furthermore, these predictions should have two key qualities:

1. The model should be specific: A model that plausibly predicts how much world population will grow in the next ten years, for example, is far more useful than a model that simply predicts an unspecified amount of growth, even while the latter model (by being more vague) is more likely to turn out to be true.
2. The model should be conditional: For example, a model that predicts a single, unchangeable rate of population growth is less useful than a model that predicts the rate of growth if governments continue with current policies, and that also predicts alternative rates of growth under feasible alternative government policies. A conditional model offers useful choices – if we do A, X will follow, while if we do B, Y will follow, so if we prefer result Y, we ought to choose to do B.

One broad class of models scores poorly on the usefulness measure: models that are fatalistic. There are lots of plausible-sounding, possibly-true ideas for how the world works that hold that your personal free will either does not exist or that if it exists, your own power to use your free will to bring meaningful change is extremely weak. Some interpretations of some religions can take a fatalistic view. Some interpretations of science are fatalistic as well, concluding that free will is an illusion, and many political theories, including most theories that almost all political power lies in the hands of some external group of conspirators, either predict an unchangeable course for the future or place all the power to change the future in the hands of a very small, unreachable elite. Fatalistic theories can also be very sticky, as they tend to excuse believers from personal responsibility to work hard to fix the problems they see.

Let us assume that we have in front of us two political theories to choose between, and after looking hard at the arguments for both, and taking into account our own biases based on how sticky the ideas appear to be, we decide that the two theories are equally likely to be true, but one theory predicts that we are helpless to change the future, while the other theory predicts that with a specific set of actions we have some chance, with hard work and a bit of luck, to tip the balance in favor of a brighter future for ourselves and our nation. The choice should be easy! If they are equally likely to be true, we might as well act on the theory that offers some hope of making a difference – if we are right, we might win a major victory, while if we are wrong, we lose nothing, since the other choice predicts that our actions don’t matter, anyway! I hold that even if a fatalistic view appears much more likely to be true than a plausible theory that offers some hope for you to make a difference in the world, you might as well go with the theory that offers you some chance to make a difference! Better still, find the most likely theory that offers a chance to make a difference!

Under this notion that a useful political model is a better choice than a fatalistic model, even when the useful political model looks somewhat less likely to be true, I argue that a certain sort of optimism, when choosing your political outlook, is actually entirely pragmatic. A fatalistic approach may be vaguely comforting, excusing you from personal responsibility to act. A fatalistic outlook might even turn out to be true. But however comforting or even true it might be, a fatalistic outlook has no chance to achieve anything useful! You might as well assume that you can make a useful difference, and that your own life, and your nation can move, perhaps slowly, but also surely, in the end, in a positive direction with your help. In my view, even failure, under such an optimistic outlook is preferable to simply accepting the fatalistic conclusion, as I would find more satisfaction, at least, in trying to make a difference than in giving up the fight before it even begins!