– or –
Why the Perfect Is the Enemy of the Good
Margaret Thatcher once said that “politics at its purest is philosophy in action”. As someone who gave her name to a political age that represented a very clear break from what came before and fundamentally changed a nation – according to both her admirers and her critics – she may have earned the right to be taken seriously.
As a self-identified member of the loosely defined liberty movement and someone professionally concerned with political psychology and persuasion, I often find myself spectating arguments between those who take a more pragmatic approach to bringing about political change and those who, although they broadly share the same philosophy, see pragmatism as an insufficient commitment to principle and, therefore, a kind of compromise of the values (such as the good of maximizing human liberty) on which their principles, and therefore their political philosophy, are based.
Within political movements, such as America’s liberty movement, this split between these two broad mindsets can become self-defeating. This can happen especially when the self-identified “non-compromisers” withhold support from their more pragmatic allies because 1) those allies don’t go far enough (such as when a libertarian refuses to vote for another libertarian because “voting is itself a compromise of the principle of individual liberty”) and 2) they spend energy arguing with their pragmatic allies that could be more effectively spent fighting their philosophical enemies.
Only this last month a dear friend and philosophical ally of mine, Jeffrey Tucker, expressed concern that what to me is a willingness to test statements of pure political principle against empirical consequences (which are determined in part by the fact that flawed humans operate in a world in which all our political principles are not perfectly realized) may make me more willing to accept an argument that contorts a principle for harmful ends.
He was thinking, specifically, I believe, about some of the worst Alt-Righters who claim to be motivated by libertarian principles but (we would agree) are motivated by anything but.
I understand Jeffrey’s concern. It’s a concern that should be felt by everyone who sees the compromises made by so many people in politics, and most dangerously by those with their hands on levers of power. It comes out of the need to hold true to what is right, even when it hurts – especially when it hurts.
And we all must do that. But when it comes to political action, that’s only half of what makes you more moral than the next man. As Orwell spent a lifetime trying to explain, it’s not what you believe that you makes you dangerous: it’s how you believe and apply it.
Perhaps because of my training as a scientist, or my ENTP personality type, I’ve felt for a long time that a desire to accept all relevant facts about the world that may affect the potential outcome of a principled action makes one’s commitment to the principle that informs that action deeper – not shallower – and one’s ability to affect the world for the good contained in the principle greater – not less.
Moreover, this seems to follow logically from the fact that political principles, of whatever stripe, exist to serve the wellbeing of people as moral subjects and objects in society. Surely, then, a principled action that is based on careful consideration of the specific psychological, emotional, and physical states of people that will affect the impact of that action because they will be affected by it, can only ensure that the action serves its justifying principle better – not worse.
Accordingly, there is an urgent need to claim the moral high ground for pragmatists over a certain kind of “methodological” or “political purist” by identifying moral and logical failures in the “methodologically pure” application of political philosophies to an imperfect world.
That is the purpose of this article, which will end by suggesting why the philosophical purists sincerely believe that they have moral high ground, when they do not.
Throughout, I shall be examining the relationship between purism (which I shall define), moral principle and pragmatism. This exploration is motivated by my own moral and principles: it is because I want to see liberty and justice restored to my country that I want my political allies to become so much more effective than we have been. And I believe that a more accurate understanding of “principle”, based on a clearer appreciation of the difference between the real world and our mental map of it is the sine qua non of our future political success.
What Is a Political Principle?
Let’s look at what a political principle actually is, and what we mean when we accuse someone of compromising one.
There is a crude sense in which political principles sit between pure philosophy on the one hand, and political action or policy on the other, linking them to each other
Broadly speaking, a principle is a general rule from which goals or behaviors that are right or morally preferable can be derived. Its justification invariably involves some moral value, such as liberty, peace, or honesty. People who hold to a principle are likely to judge a behavior or state of affairs as better or worse inasmuch as it manifests more or less of the moral value of interest.
Again broadly speaking, a political philosophy, such as libertarianism or communism, is a logically related set of truth claims about the nature of the world or society. Different philosophies depend on different fundamental precepts, definitions and values. For example, for a communist, “class” is a fundamental precept. “Equality” may be a value. For a libertarian, “natural rights” and “aggression” may be fundamental precepts and “liberty” a fundamental value. Moreover, different philosophies are typically concerned with maximizing different moral values or minimizing different moral harms. For example, Sharia Law seeks to maximize obedience with a religion; libertarianism seeks to maximize personal freedom; a certain form of progressivism seeks to minimize material hardship. Of course, adherents to a philosophy may well argue about the exact meanings of these fundamental precepts and values, and how they apply in the real world, but the point is that a commitment to a political philosophy involves a commitment to both values against which human experiences are measured and principles that translate those values into rules for behavior.
I used the term “human experiences” there, which is critical. Politics and morality have human beings as both their subject and object. Necessarily, political principles are related to the lived experiences of people; in a nutshell, “the only legitimate ends of politics are people”.
It should also be clear, then, that even though we often rightly say that we do something “simply because it is the right thing to do”, ultimately the rightness of the action has something to do with some kind of human experience. The relationship may be indirect, may only become obvious in the long-run, and may depend on unstated assumptions about aspects of human nature such as reciprocity, for example, but still the rightness or otherwise of a principled action is inherently related to human experience.
That statement is perhaps so obvious as to seem almost banal.
But it is too often forgotten in politics by those who test their actions only against their axioms. Moreover, it is critical because it reminds us that the only possible data against which we can ultimately test whether a) a claimed principle in fact serves a value, or b) an action serves a principle, are data about human experiences that may be caused or affected by an action.
Political actions and policies that anyone may spend any time arguing for or working toward are, by definition, specific expressions of principles that seek to act on the world as it is – rather than on a hypothetical, abstract or idealized world – to increase or maximize a value or reduce or minimize a harm. What is perceived as a value or a harm to be increased or reduced typically varies with the associated political philosophy.
Those performing the principled action may do so because they expect it to have an immediate effect that clearly favors a value, or because it is consistent with a principle they hold – even if the specifics of how the action will promote the favored value cannot be exactly articulated or known. In the latter case, the specific human effects may not be predicted, but the commitment to the principle that justifies the action rests on a commitment to the value that justifies the principle in the first place.
The point that political action (like all action) operates (out of logical necessity) in the world as it is, is crucial because a political philosophy may very precisely tell you what an ideal world might looks like but almost never tells you the best way – even measured in terms the core values of the philosophy – of getting from the current state of affairs (which have arisen out of historical contingencies) to the ideal state of affairs – except inasmuch as I will describe below. In other words, values and principles may give you your ultimate political or moral destination. They may even give you a broad direction. But the intermediate outcomes along the “route” from the non-ideal here-and-now to the ideal there-and-then can only be determined as we go – and their goodness, measured in terms of the core values of the philosophy, can therefore also only be known as we go.
The Map Is Not the Territory. Your Theory is Not the World
The question of which actions are principled is not trivial, precisely because principles, the moral or political philosophies with which they are associated, and the values that they favor, are necessarily formulated in abstractions and generalities – not explicit descriptions of individual human experiences, which are their ultimate target.
This raises at least two problems. With respect to abstractions (such as “liberty”), people can reasonably disagree about how they map to specific human experiences. When it comes to the generalities (such as “don’t initiate aggression”), people can reasonably disagree about how, or the extent to which, they apply in specific cases. Both problems can often be regarded at some level as a disagreement about the meanings of words.
Critically, principles that are formulated in terms of generalities and abstractions do not specify the limits of their own application. They generally do not provide “meta-principles” that can be applied to resolve conflicts that arise when two principles seem to suggest different courses of action in a particular situation. Nor do they provide ways to measure one value against another (life vs. liberty) when different values that justify different principles favor different actions.
It is easy to posit real-world hypotheticals that make even some the most widely accepted principles (the golden rules) non-trivial in application.
For example, the Hippocratic principle, “first, do no harm” can be seen to be problematic by considering one of my favorite examples of a man dying on the street after being hit by a vehicle. Imagine that a doctor on the scene could save him but refuses to do so for lack of payment. You don’t have the means on you to pay the upfront price he is demanding. Against the clock, as the man is suffering from shock and bleeding to death, the only way you can see of saving his life is to break into a nearby house and steal the valuable-looking piece of art in the window to pay the doctor. (You can settle it up with the owner after the fact.)
A methodologically purist reading of the Hippocratic principle would require that you do not break into the house (because it actively harms another) and so stand by and watch the doctor do nothing as the man dies. A second adherent to the Hippocratic principle could interpret it differently, by saying this is a damned-if-you-do, damned-if-you-don’t situation. Since the bystander would be doing harm either way – letting the man die or breaking into the house to steal – he should do the latter to pay the doctor, and then put it right with the homeowner to minimize the harm.
Yet a third adherent to the principle might agree that the first response is indeed the methodologically pure application of the principle, but since in this instance, it leads to a morally unacceptable outcome, we have learned something about the limits of the application of the principle. This is the most interesting position, and probably the most accurate. What are those limits of its application? They seem to exist because the word “harm” is not exhaustively and precisely defined, or at least it is not defined in the same way by all who may be interpreting it.
The problem of applying this principle in this example more intriguingly has something to do with the degree of moral responsibility that comes with actively causing harm (stealing) vs. that of allowing (an easily prevented, in this case) harm (dying man) to occur. On the one hand allowing harm you didn’t cause isn’t “as bad” as causing a harm, but on the other hand, our same moral intuition that informs “first, do no harm” tells us that to be left to die unnecessarily is worse than having something stolen. But the principle itself does not tell us how to weigh one against the other… so we need to refer to the value that the principle seeks to maximize, or a meta-principle, to enable us to determine how to apply the Hippocratic principle in this case.
For another example, how about the golden rule, “do unto others as you would have done unto you”? That becomes problematic as soon as you allow for the biological, emotional, or just experiential and cultural differences between people – since how you may want to be treated or spoken to would have an entirely different effect on someone else. So does the golden rule mean you should actually do the behavior that you’d want done to you, even though it may have a very negative impact on the other person, or does it mean that you should strive for a person you affect to experience the same outcome that you’d want to experience, which may require you to behave very differently toward them from how they would need to behave toward you, to get the same result?!
As a concrete example, which is in my mind because of the recent #MeToo phenomenon, many men would be very happy to receive a picture of an intimate picture of a woman whom they are communicating with online but have never met, whereas most women are repelled by receiving the same from a man. One might say that a methodologically pure reading of the golden principle would entirely justify a man’s sending an intimate picture to a woman he doesn’t know, but to do so isn’t actually principled by any solid moral intuition (serving the value that the principle seeks to maximize). Why not? Because the principle, which is necessarily abstract and general, can only be manifest in specific cases, and only consideration of the specific outcome in this case reveals that the value that principle exists to serve (increasing human happiness, or at least never reducing it) is not served by its “pure” application in this case. Again, we see how that the principle logically cannot contain the metaprinciple that governs its application in marginal or complex cases.
It’s already getting tricky and we’ve not even gotten to the specific logical errors of methodological purism. But we shall now.
The Problem of Purism
We all know methodological purists, as I shall call them. These are fervent people who experience themselves as highly principled. They are characterized by preferring to implement the policies that would prevail in their ideal world as fast and completely as possible in the real world, without consideration of the particulars of the current state of affairs or the situation or preferences of the living people (as opposed to hypothetical people in a hypothetical future without distinguishing experiences, which political philosophies more naturally deal with), whose lives would be affected by the change. As a result, they often refuse to do anything they would not do in their ideal world (like vote) and they refuse to support policies that, although favoring one of the core values of their philosophy (increased liberty, reduced government interference etc.) do not “go far enough”, because they would allow the perpetuation of certain non-idealities: were the methodological purist to support the new policy, he would experience himself as actively endorsing not only the move in the right direction but the limited degree of the change. Their refusal to do the latter is often, quite reasonably, why they experience their positions as more principled than those of their philosophical allies who are happy to move the ball down the field in the right direction even if that means agreeing in some formal sense to stop before the end zone.
As a result, they tend to see active support for policies that are not philosophically ideal as a compromise of principle, even if there is good reason to believe they advance the values that justify the principles that are consistent with their philosophy.
I call this, “methodological purism”. Its problem is that while its adherents experience themselves as “acting on principle”, their actions are in many cases less principled once we realize that the word “principle” doesn’t refer only to the state of mind of a person who acts, but also, and more importantly, to the effects that an action have on the world.
Specifically, to act properly from principle requires an active engagement with the contingent, empirical facts on the ground – the here-and-now – to determine whether the supposedly principled action at a particular time and place (such as voting for a candidate, supporting a policy, opening a border, redistributing wealth) will have effects that increase the very value that the principle exists to promote.
Failure to engage in that way will usually cause a methodologically pure action in pursuit of a principle to have outcomes that disserve the very value that makes the principle a principle in first place.
Specifically, methodological purists make three fundamental errors that cause them to act in ways that feel principled (philosophically pure) but are not (in terms of their outcomes). The first two errors concern the world; the third concerns their knowledge of it.
- They fail to consider the transient effects of principled action, and to test those effects against the values on which the principle that justifies their action depends.
- They fail to consider the “problem of second best”, which is as follows. For a complex system, (such as a society), although an ideal outcome (as measured in terms of any value of interest, such as “liberty”) requires particular settings of relevant inputs/variables/policies (e.g. welfare state = 0, taxation = voluntary, borders = open), when any one of those inputs/variables/policies is fixed away from its ideal setting or value, then generating the outcome that is as close to the ideal as possible generally requires that some of the controllable inputs/variables/policies are set well away from their “ideal” values.
To summarize the above, the outcomes of a complex system depend not only on the variables that govern or describe it, but also on interactions among those variables, so knowing everything about the ideal state of the system doesn’t necessarily tell you anything useful about how to get from a non-ideal state to a closer-to-ideal state.
- Methodological purists mistake the precision and/or logical consistency of their political beliefs for the accuracy of those beliefs. In fact, when knowledge is imperfect, its accuracy is likely to increase with imprecision and inconsistency!
Since the above is abstract, here’s a concrete example for our time. Considering libertarianism again, if the ideal world that maximizes liberty involves open borders (variable: border constraints = 0) , then the most moral policy for a methodologically pure libertarian is to open borders as much as possible as fast as possible, regardless of the current state or any other variables and, therefore, without analyzing the chain of cause and effects were we to do so.
This approach to politics makes two logical critical errors of omission that arise from the fact that a community is both a complex system and a dynamic one, in which not all variables can be known – let alone controlled. 1) A community or country is a dynamic system and so policy changes have transient effects. The transient effects are not controlled by the same variables as control the ultimate or equilibrium state of the system. 2) A community or country is a highly complex system that is governed by more variables than can be known or controlled and those variables interact with each other in ways that cannot be determined even from exact knowledge of its ideal state.
We can distill the problem to what I shall call the “methodological fallacy”, which is that the closer all variables are to their ideal values, the closer the outcome will be to the ideal. The assumption is fallacious for any system in which 1) some variables are held by external forces at their non-ideal values, and 2) the current state is a function not just of the current values of the variables, but by the path by which they arrived at those values. Even the simplest society is such a system.
The world is highly complex. When it comes to doing politics, we can only perturb a particular part of it at a time (we can change only some of the variables/policies that govern it) or we can perturb it only slightly (we can change the variables/policies that govern it only by small degrees). In other words, we can neither get to our ideal state (whatever your philosophy may be) in one shot, nor can we predict all the consequences (intended and unintended) of the (partial) change (intentionally for the better) that we may cause as we try to get there.
For methodological purism to achieve the outcomes its seeks to achieve, it would be necessary to control all variables in a society simultaneously, and for their effects to be both instantaneous and predictable – conditions that never apply in the real world – not even in the world of objects, let alone that of people with free will.
Society Is Dynamic: Transient Effects Matter.
Political philosophies tend to focus on characteristics that should prevail or be absent in an ideal, or at least targeted, state, which its adherents seek to bring about. In this ideal or target state of society or the economy, important various policies or institutions will be in effect and variables will have different settings (often zero or as high as possible). For libertarians, the variables of principle may be the size of government, amount of aggression against others; for communists, they might be the amount of private ownership or degree of political control of assets by workers; for progressives, they might be the gini coefficient, or the number of people below a specified poverty line.
As an example, in an anarcho-capitalist’s ideal equilibrium society, the static input variables might be amount of government = 0, protection of private property =100%, etc. maximizing output variables of maximum liberty and justice. Of course, anarcho-capitalists may define those terms differently from those with other political views, but the important thing is that these are the metrics, however defined, that matter to them.
Here’s the critical point: the variables that are required to fully specify the favored, or even ideal, equilibrium state or direction of society, do not control and cannot be used to predict the response of a society to a change in any of the prevailing policies – especially in the short and medium terms (which is when those who currently comprise the society are alive).
To understand how transient effects in a dynamic system sever the link between methodological purism and principle, let us consider one of the simplest possible dynamic systems in the physical world.
Imagine you have a simple electrical circuit with the three basic electrical components in it – a capacitor, an inductor and a resistor. (It doesn’t matter if you don’t know what these are.) Their respective properties – capacitance, inductance and resistance – are all set to various values. When a particular input voltage is applied to the circuit, the circuit generates a corresponding output voltage, which depends on the input voltage and the properties of the components. If the components are all fixed, then you can control the output voltage by controlling the input voltage.
That’s simple enough, and it’s possible to plot a curve for the circuit that shows what output voltage is generated from any input voltage for a particular version of the circuit.
Now, this simple circuit – with one simple input, a few basic components and one output of interest – is an analogy to a much more complex, much more dynamic, much less predictable (because of millions of people with free will etc.) society. The input to the circuit and properties of the components are analogous to anything that can vary in social, economic or moral life, such as government revenue or tax policy, and the output could be a social, economic or even political value, which could be something concrete, like the number of homeless people in the country, or something abstract, like “liberty”. Obviously, the link between government revenue and homelessness or the details of the tax code and liberty, prosperity or wealth distribution is much, much more complicated than the link between the input and output voltages in the circuit example, but although the analogy is infinitely simplified, it suffices to show that, even in principle, suddenly changing an input variable to its ideal value – which is the value it would have in an ideal world with the ideal output value at long-run equilibrium – does not yield that best outcome before the ideal equilibrium world is reached. And that’s before we even consider the fact that more complex systems (still much less complex than a country) do not in general even have a long-term equilibrium in which a particular value of interest is stably constant, increasing or decreasing!
What happens between the initial state (before the perturbation of the system, such as occurs by changing the input voltage of the circuit or opening the borders of a country) and the final state (after the perturbation) is determined by the initial state of the system as well as the properties of the components of the system that are of no interest to someone who is seeking to shift the system toward an ideal state perhaps because they do not, in principle, determine anything significant about the system in its ideal/target state or perhaps because they are not even recognized as controllable components.
In other words, although you may be able to determine from your philosophy that in the ideal world, the inputs (policies) would be those that you seek to set based on the principles consistent with your philosophy, you cannot determine from your philosophy that their ideal settings would deliver the outputs you seek before the ideal end state is arrived at.
Rather, other factors must be considered to determine whether a transition that will be caused by a change in policy is consistent with the principles that justify that change.
A diagram helps to clarify the point.
A B
In both graphs, a system is initially in an undesired state (initial value) and we want to move it to the desired state. The vertical axis represents the metric that matters. It could be the output voltage of the aforementioned circuit, but in a country, it could be “liberty”, “prosperity”, or “justice”, however defined. Sure, it’s unusual to put such a big concept on an axis – but this is a visual analogy, and if we are going to use any such concept to say one state of affairs is better than another, then it is completely consistent to represent it in this way.
In both A and B above, we perturb the system (change the input voltage of the circuit or the immigration policy of the country) to whatever values will give us our desired end state. The very attempt assumes, as do most people and certainly political purists, that this is even theoretically possible. If it isn’t, the methodological purism ends there. But let’s assume it is to give purism a chance.
In case A, the input variable (input voltage, immigration policy, welfare state) is changed to its “optimal” value (input voltage goes to 1V, borders are opened, federal welfare is abolished), but the immediate effects of the change have nothing to do with what the end state in the ideal world with the input variables are all as desired. Rather, in both a circuit and a country, we can get huge transition effects (a massive spike in the output voltage, increased support for authoritarianism, millions of starving children etc.) that are at least as far from the desired end state as the initial state with respect to the very value or values by which the final state is preferred to the initial one. This occurs because we are moving between states, and how something moves is a function of a) where it currently is and b) many variables that didn’t need to be considered to predict the final, long-run equilibrium state.
In case B, the situation is very different: progress to the end state is smooth, with a consistent approach of the metric that matters (output voltage, liberty, prosperity) to its ideal end value.
Now this is a critical point: it is much harder to get a smooth transition like that in case B than it is to get the wild disruptions in case A. The transition in B requires more variables to be adjusted more carefully and slowly than the single change in one variable that yields case A. And those complex changes that are required to generate the smooth transition in case B cannot be determined even from a perfect understanding of either the ideal state or even the gap between the current state and the ideal state. Factors that don’t even affect the end state but do affect the transition have to be explicitly identified and controlled.
In other words, to get to the principled outcome as reliably as possible, the “methodologically pure” move (change the input voltage instantaneously, open the borders, abolish the state tomorrow) won’t work. In short, methodological purity does not translate into “outcome purity” in a dynamic system so “pure actions” don’t get “pure results”. Therefore, inasmuch as principles exist to promote a political or a moral value, “pure actions” aren’t the most principled.
In politics, people who are getting hurt (socially, economically etc.) in a transition resist that transition. And they should – because the methodological purists, also called ideologues, don’t much care about them. They are the people whose suffering is “justified” by the philosophy.
It is their sacrifice that explains, for example, why even “libertarianism”, a philosophy whose core value is not to impose on others, can indeed be “imposed” just like any other political philosophy, even though “imposing libertarianism” should be an oxymoron. While it is true that those hurt in a transition may not be being hurt because flaws in the philosophy itself, they are nevertheless victims of the fact that the philosophy is being imposed by a sudden change in a policy (abolish the welfare state tomorrow, open the borders immediately) in a methodologically purist fashion such that the “delta” (change) between the old policy and the new policy has negative human effects. If “increase liberty” is the justification for such a change, then the change felt by those negatively affected individuals in the transition will be at odds with the purpose of the change.
The argument works for any metric – not just “liberty” – and for any philosophy. For example, imagine we’re going for increased “prosperity”. A new tax system may likely be massively better for the country than the old one in the long-run, but the change may cause a significant reduction of prosperity to the whole country for an indeterminate period and to some people for ever.
The general point is that by the principle that justifies a political change, the change is itself harmful to some and potentially all for an unknown period. A more principled – but less methodologically pure – approach to the change would be to reduce the harm done in the transition to the better end state, often by reducing the transition period, and that often by reducing political resistance.
The problem becomes greater as the policy change becomes greater and our ability to predict the human impacts of the change declines.
To keep with the immigration example, consider this. Recent research on psychological and political authoritarianism is extensive. Citizens with an authoritarian tendency – who are most inclined to use force, including the force of the state, to impose their own view of society on everyone else – tend to come to the fore, gain power and support policies that limit the freedom of others – under conditions of “normative threat”. This threat is the perception (right or wrong) that the range of cultural, moral and political norms exhibited by people in a community has become too diverse and even mutually conflicting to enable that community to remain stable and to enable them to retain their identification with that community without suffering negative consequences.
Those same academic studies – not to mention recent news from Europe – show that these conditions of normative threat, which lead to the political ascendancy of authoritarianism and exclusion, are (obviously, by the above definition) exacerbated most seriously by a large influx of people who do not share their adoptive community’s cultural, moral and political norms – and most grievously when those people do not assimilate. The larger and more sudden an influx of immigrants, the less the assimilation, and the greater the authoritarian reaction. In other words, a methodologically pure action to increase liberty in principle – opening the borders – is highly “outcome-impure” because it reduces liberty in fact. Whereas the sudden opening of the border may have been a “pure libertarian” move in a narrow methodological sense, it wasn’t a principled move once the outcome is considered and the valued advanced by the principle is “liberty”.
Considering these transition effects, the principled libertarian should indeed move toward the open border ideal, but in a controlled fashion that accounts for the self-defeating transient effects.
It’s worth going one step further with this. As Jordan Peterson recently explained concerning this exact matter:
“a complex system cannot tolerate extensive transformation over too short a period… it should not be assumed that citizens of countries that have not evolved functional individual rights-predicated polities will hold values in keeping with such polities.”
And you don’t need to consider a system as complex as a country to see it. Many much simpler physical systems exhibit this truth.
The problem goes even deeper than that, though. As a clinical psychologist concerned with individuals, Peterson understands that the changes to our culture, such as those caused by uncontrolled immigration, affect not just the broad characteristics of our society and polity, but also the very ability of individuals to function in the most basic ways. In other words, large-scale cultural change has a direct effect on individuals, such that a reasonable expression of individual liberty would be to seek to manage that effect in the least harmful way.
Your culture is a set of value-laden presuppositions that you [use to] orient yourself in the world that match the set of value-laden presuppositions that everyone in your culture has and acts out. And so what that means is, … there’s a match between what you see and what you’re doing and what other people expect, and it’s that match that regulates your emotions. It’s not the belief system: it’s the match. And so part of the reason why people are so tied to their cultural identity is because their cultural identity regulates their emotions, and in a profound way… One of the things that stabilizes human nervous systems is… [t]he societal structure [that] regulates your emotions because of the match between your expectations and the behaviors of the people within that structure.
I quoted Peterson’s explanation of the relationship between individual emotional stability and the individual’s cultural context because, for me at least, it further provides something of a principled basis for not rejecting the above finding – about mass immigration causing an authoritarian reaction – with a simplistic, “you can’t make an omelet without breaking eggs” attitude or a kind of religious belief in open borders as a means of increasing the sum of a society’s liberty or even the liberty of humankind.
I went into that detail on immigration not to take a position on the issue of open borders but to indicate how variables other than the ones that libertarian open-border advocates think matter in assessing the relationship between borders and individual liberty actually matter a great deal. Despite the apparent moral simplicity of the borders issue in libertarian theory, limitations on our understanding of each other and our world mean that we typically cannot predict the outcomes of a principled policy change we support, even in terms of the value that justifies and informs the principle (liberty in the immigration case).
So here’s a moral rule: our ignorance demands that we test the outcomes of our principled actions against those values that justify those principles as the only way of determining whether those actions truly were the most principled.
And here’s a practical rule: the complexity and dynamism of the world in which we operate is such that a politically pure method turns out not to be the most effective way to move toward politically favored outcomes.
A hardcore purist may still object to the above with something like this: “It doesn’t actually matter what happens when the borders are opened. Open borders per se are morally superior, regardless of outcome – just as it was right to free the slaves, regardless of outcome.”
This is a great example for testing the thesis because ending slavery was so obviously right and necessary that no one would argue that it shouldn’t have been done – and as quickly as possible.
And yet. And yet. The Brits took a different approach to eliminating slavery from that of the Americans – one that was in an important way methodologically impure. A modern-day comparison of the two countries’ race relations shows overwhelmingly that the methods used by the two countries made and continue to make a huge difference to the outcome for the liberty of those affected by slavery – even though both nations clearly moved from an initial state with slavery to a target state without slavery.
From a libertarian perspective (and most other political perspectives), the methodologically and philosophically pure way to free slaves was just to free them, by force if necessary, because a human being cannot be held as property by another. It follows that the slave-holders had no right to compensation for loss of property when they lost their slaves.
However, if you’re starting in a world in which the evil of slavery exists and, moreover, many influential people – perhaps collected into large states with armies – believe that they owned their slaves fair and square and they had a right to do so, then a concession to pragmatism, while morally impure, might well be principled. Why? Because if you pay slavers for their slaves (despite the apparent immorality of it) and then release those slaves, then you may achieve a much faster and easier transition to the desired end state (of maximal liberty enjoyed by everyone in the country) – and most importantly, the liberty of the freed slaves and their immediate families may transition much sooner to its proper value. If, on the other hand, you don’t pay the slavers off but remain morally pure by imposing your philosophy with morally justified force on the slave owners, then those slavers will likely feel that something has been taken from them by force, and they may end up starting a civil war that takes hundreds of thousands of lives. Since a dead man has no liberty, the short-term outcome was opposed to the most fundamental value (liberty) that justified the change to the end state in the first place. Since the act of force provided fuel for the resentment of the moral and military losers that lasted centuries, the long-term outcome also fell short in terms of what could have been accomplished for liberty, justice and prosperity.
In short, the pragmatic British solution (pay the slavers) resulted in less racial tension even centuries later, such that British citizens who identify today with enslaved communities of the past feel nothing like the inability to participate fully in their nation’s liberty that is experienced by African-Americans.
For another example, consider the astonishing peace process in Northern Ireland. The conflict between Irish Republicans and the British state has gone on for centuries. Both sides hold simple principled positions, which are mutually incompatible.
The Irish Republican position is that the north of the island of Ireland was occupied by a foreign power that had no right to be there. Therefore, the philosophically and methodologically pure and consistent thing to do is fight, making no concession to the invading power. The principles on which their actions were based served the Republican values of wellbeing for Irish-identified Catholics in Northern Ireland and the sovereignty of the island of Ireland.
The British position is that the north of the island is in fact a British sovereign territory and the government has a duty to its citizens there to protect them. Therefore, the philosophically and methodologically pure and consistent thing to do is to govern, and make no concession to the terrorists who were killing innocent citizens (which they were). The principles on which their actions were based served the British values of wellbeing for British citizens in Northern Ireland and the sovereignty of the United Kingdom.
Yet, the solution to this centuries-old impasse was a recognition that the conflicting values of both sides that justified their actions against the other utterly principled, could actually only be served by pragmatism.
The British government negotiated with the Republican terrorists, who, to them, had no moral or legal legitimacy whatsoever – a methodologically and philosophically impure move, while the Republicans in turn, negotiated with the invaders, who, to them, had no moral or legal legitimacy whatsoever – a methodologically and philosophically impure move. The agreement they came to, involving a new assembly to govern Northern Ireland, served the values that drove the principles of both sides, making the agreement itself a principled one for both parties but a methodologically pure outcome for neither.
The peace brought about by the establishment of the assembly advanced the wellbeing of the British citizens in Northern Ireland. Simultaneously, the Irish Republicans gained formal recognition as a political power in that same assembly and, therefore, for their position of eliminating the British from Northern Ireland. Both sides achieved a principled advance, in the direction of the values that justify their principles – but they did so through pragmatism – putting improved outcomes over methodological purity.
Hundreds, perhaps thousands, of lives have been saved as a result. (Does that serve an even deeper principle?)
The general point here is that whereas principled behaviors put people first by considering actual human outcomes, methodological purism is about putting a philosophy first by considering consistency between a particular action and a believed abstraction about the world, absent real-world context.
If you still doubt it, here’s a thought experiment.
You are a libertarian and you can push the button today to eliminate all government spending on everything – medicare, Medicaid, pensions, welfare… you get the idea. But if you push it, it all goes away right now. The machines in the hospitals that are paid for by government money stop. The welfare recipients receive nothing more. The borders open and you have to welcome everyone who comes into the States. The population may double in a couple of months, and the majority of it may vote against common law, against natural rights and for an authoritarian, big government.
There’s no time to prepare for any of it. There’s no time to put in place any of the private solutions to society’s problems that would exist in the libertarian’s ideal world… and indeed, could exist in our future, if we didn’t press that button but made changes slowly enough for those solutions to develop, establish themselves and become the norm.
If you’re prepared to push that button, regardless of any of the things that could happen over a time frame that you cannot know, and you don’t feel that your libertarian principles (favoring life, liberty and property, for example), require you to find another means to your ends than pushing that button, then congratulations: you are a methodological purist. If your inability to make empirical predictions of what would happen when you push it doesn’t cause you to hesitate – not in spite of your love of liberty, but because of it – then you’re a methodological purist, and you’d probably feel that you’d be more principled than those of your philosophical allies who wouldn’t push it.
But I believe you’d be wrong.
Because it would be your more hesitant and pragmatic friends who would recognize that, while agreeing with you that the ideal world would have in it all those changes with voluntary organizations and freedom of movement etc. etc., the real outcome of the change from here to there in terms of liberty and prosperity would likely be so negative that the methodologically libertarian act of pushing that button would lead to an “outcome-impure” world of less liberty, prosperity and peace.
And in that concern, your pragmatic friend would be the principled one. He would rather have a more complex button – or series of buttons – that wind down the current state of affairs in a way that does the least harm (in terms of our agreed upon libertarian values) to people (the only legitimate ends of any political philosophy) over time.
And if you are concerned by the weakness of your commitment to your own beliefs because of your discomfort about pushing that button, don’t be: that discomfort just means that you instinctively know what science tells us – that the a change from A to B cannot be fully understood by understanding simply the difference between A and B, because the things that happen between A and B are determined by factors you can’t control and don’t even know.
Society is Complex: You Can’t Control All the Variables that Matter.
As we’ve discussed, principles concern the experiences and behaviors of human beings. The impact of an action on people is determined not only by the action itself but also by context, which includes all of the other variables that affect that describe the system. For example, quitting your job tomorrow is not an inherently principled act. But if you have just been asked to participate in something immoral that is being done by your company, then it may well be.
We have seen above how methodological purism ignores the transient effects of actions (which are important because society is dynamic). Here we will consider the second problem: methodological purism also ignores their context, which is hugely important in a highly complex system.
The fallacy is in the assumption that an ideal outcome requires that a set of variables (policies) take certain values, then setting any one of those variables to that value is always better than not doing so, regardless of the settings of the other variables.
So here’s the other critical point missed by the methodological purists: The effects of variables you do control depend on variables you can’t control. In other words, the outcomes of a system are caused not just by the input variables, but also by the interactions among those inputs. Almost any example of a system that exhibits the transient effects described above, including those already presented, also exhibits this interdependence of variables.
Consider the conundrum of nuclear disarmament. The moral value to be maximized is “peace”; alternatively, the moral metric to be minimized is the number of human deaths or the probability of wiping out humanity. Clearly, the ideal equilibrium world would be one without weapons of mass destruction.
To a methodological purist, since in the ideal end state, the holding of weapons of mass destruction of each state equals zero, this would justify the unilateral disarmament of one’s own country. But this would not yield the closest-to-best-possible outcome given the variables that can’t be controlled. In this example, it is easy to see that if an enemy country maintains such weapons, then the most principled action in pursuit of peace is not to disarm but is to remain armed.
This is of course what happened in the cold war, as part of a strategy of “mutually assured destruction” (MAD). The general point is that the context (uncontrollable variables) changes the action that is most principled from the “methodologically pure” action to one that is methodologically impure but delivers the best possible outcome in practice.
Readers who know game theory will know the MAD strategy as a Nash equilibrium: such an equilibrium is sub-optimal for all involved if hypothetical worlds are considered, but represents the optimal choices for each party given the current state of the real world.
The most general form of this insight is critical to any understanding of what principled behavior means and has been called the “theory of second best”.
The theory of the second best concerns a situation in which one or more optimality conditions cannot be satisfied. If one optimality condition (my enemy has no WMDs, the welfare state cannot be abolished immediately) cannot be satisfied, then it is possible that the next-best solution involves setting other variables away from the values that would otherwise be optimal (from having no WMDs to having WMDs, or from open borders to controlled immigration).
It is relatively easy to construct exact mathematical or quantitative examples. For libertarians, for example, the ideal world has open borders and no welfare state (both maximizing individual private property). But once you have a welfare state, if you can’t get rid of it, then inasmuch as open borders are likely to expand it by increasing welfare roles, taxes, government spending, native resentment of immigrants, then a calculus of goodness of outcomes based on the value of liberty, is clearly affected.
Since it is easy to see that “methodological purism” is not the same as “outcome purism”, why does pragmatism that implicitly recognizes this fact feel like moral concession to the more ideological among us?
The answer lies in the third fundamental error of methodological purism. It is an error not about the world but about our knowledge of it. In other words, it’s an epistemological error.
When Data Are Imperfect, Accuracy Is Not Found in Consistency, Certainty or Precision
Why do methodological purists (“you’re not a principled libertarian if you don’t want completely open borders”) tend to experience themselves as being the more principled than those who seek to manage political transitions in ways that bear in mind the specific experiences state of, and transitionary difficulties that may be faced by, those affected by policy changes?
The answer involves an epistemic illusion.
Let’s say that A and B are two related empirical facts about the world. Like almost anything in the real world, these two facts are related to many other facts. Once again, let’s model the situation as simply as we can, and treat A and B as simple numerical variables.
Now, imagine that you and I are both trying to build an accurate and coherent view of the world based on our observations of it. And let us imagine that, regarding these two facts that so interest us both, we agree for whatever reason that A+B = 10.
We go out into the world to measure the values of these variables, but our ability to do so is limited – because we can only measure them indirectly, or perhaps they represent something really complex and we have no exact instruments to measure them. The reason for the limitation is unimportant.
In any case, I observe A to be somewhere between 3.8 and 4.2 and I measure B somewhere between 4.3 and 5.7. I know my instruments are imperfect, and I’m an imperfect user of them so I tell you that my best guess at A is 4 and my best guess at B is 5, but really I’m not going to insist on my own rightness or anything tighter than those ranges.
You laugh at me because you know I’m wrong and you know that I know that I am wrong – because 4+5=9 – not ten. And even if I go to the end of my ranges, 4.2+5.5=9.9, I still don’t get to ten. So by any interpretation, I’m still wrong because there is no way that my conclusions can possibly represent the world accurately. I agree with all of that, but say that given my limitations and the tools available to me, that’s the best I could do. I know I’m wrong somewhere, but I have good reason to believe I’m less wrong than most because I’ve spent a lot of time on this and have been genuinely concerned with getting the best answer I can. I am probably motivated to make a closer study of A and B to improve my estimates and I admit that I’ll never get to the right answer.
Now, like me, you’ve gone out and measured A as well, except that you came back with a firm A =4 and since you know that A+B=10, you deduced B must be a firm 6. And you tell me that, whereas my understanding of the world must be wrong, yours could well be right. And it could – but because A=4 and B = 6 give you two critical things that my estimates don’t have – consistency and precision. And those are indeed very important characteristics of a view of the world, because it is indisputably true that a completely accurate view of the world must exhibit both of them. You feel good to have your understanding of the world hang together like that.
You get have both consistency and the precision that come with 4+6=10. Combined, they give you something else that I don’t have – a feeling of certainty.
Moreover, your logical consistency makes it a lot easier for you to make claims about all the other things (C, D and E etc.) that depend on both A and B. Unfortunately for me, perhaps, my view of the world (C, D and E etc.) gets very tentative very quickly, because I’m trying to account for my initial imprecision and my knowledge that I was wrong out of the gate. Meanwhile, you have confidence and a beautiful coherent set of beliefs about A, B, C, D and E. Wow. The more you build your system, the more of the world you are able to explain, and it feels better and better.
But it turns out that although I was less precise, less logically consistent, and less certain, I was the more accurate of the two of us.
You see, it turns out that the true values of A and B were 4.2 and 5.8. The sum of the errors between reality and what my knowledge was 0.1. The sum of yours was 0.4.
That’s the epistemic illusion of the purist. And it’s worth bearing in mind that neurologists now understand that the certainty of knowing has very little to do with being right.
In a world where we cannot know with precision everything that matters, the comfort the feeling of knowledge that comes with precision and consistency, and our mistaking that mental state as evidence for the accuracy of our picture of the world, is but a beautiful siren that calls us from the scarier, less certain, open waters of accuracy and truth.
For human beings in a world they cannot entirely understand, a paradigm that cannot possibly be entirely right but admits its own inaccuracy and imprecision is likely to be more accurate for that very reason, than one that asserts an accuracy and precision that both our limited ability to experience the world and the approximations that are built into the concepts that we use to systematize it, cannot possibly justify.
My map of the world is derived by taking more measurements of the world and doing the best I can with the data. The purist marks a few landmarks that he is certain about and then locates other things in the world in exact, precise, and consistent relationships with those landmarks, of which he is quite certain. But, if he fails to test his map against the territory whenever he uses it, then he cannot discover the increasing divergence of his actual path from his intended path. His map is the more beautiful and may even be perfectly right in a few places – but the pragmatist’s is more accurate overall: it’s a bit wrong everywhere, but it is more useful in getting to where I need to go (rather than just believing I am) because the legend on it reminds me to take care of the margin of error when I’m navigating.
Getting the Balance Right
Nothing in this article is not an argument for moral fluidity. It’s an argument, I hope, for real principle action and other-people-centered intellectual humility.
Yogi Berra said, “in theory, there’s no difference between theory and practice. In practice there is”.
I believe that there is a sublime truth that would serve well anyone who would change the world for the better.
The world that we must change through our principled actions is the same world that provides the feedback against which we can check the goodness of our principles not as measured by the rest of what we believe, but in their application to the lives of others.
And when the thing acted upon is necessarily the only thing that can ever inform our action, we have a problem of self-reference and feedback. A system that is characterized by such things never affords simple, static, answers to the questions that may be asked of it. So let’s not pretend it does.
I started this article with Thatcher, so I’ll end with her. She was a conviction politician of an unusually pure type. Her conviction and principles enabled her to change the country of my birth. And, I suspect like my friend Jeffrey, whom I also mentioned earlier, I know that the world will bribe the principled person with the convenience and material benefits that come with compromising them. The moral values that they exist to serve require us not to be to do so – except where the compromise is itself principled or, perhaps more accurately, meta-principled.
Might a thesis like this one be dangerous in the hands of those without moral fiber who would like to dress up their moral selling out as good sense? Only if they utterly ignore the insistence on insisting on the morality of the outcomes we cause. Moreover, most of us, using those hard-to-define but impossible-safely-to-ignore things called gut and conscience, know the difference between making an excuse for a morally compromising action and doing the best we can when the contingencies of the world prevent us from getting an ideal outcome in one area of another – as they almost always do. And that’s why we must understand our principles in such a way that we apply them deliver us the best possible outcomes – so that the perfect is never allowed to become the enemy of the good.
Those things – gut and conscience – as well as intelligence and cold rationality and logic, applied to human actions and experiences, provide us with the moral intuitions about specific situations that are ultimately the only data against which we can test our moral, and therefore political, principles in practice. And logical consistency requires that we do so, because such moral intuitions are, ultimately, the only data that can ever justify any political belief in the first place.