...Archive for July 2019

The externality lens

Will Wilkinson has written a provocative paper that tries to explain a remarkable regularity in American politics. Wilkinson writes:

[T]here is now no such thing as a Republican city. “As you go from the center of cities out through the suburbs and into rural areas, you traverse in a linear fashion from Democratic to Republican places,” Stanford political scientist Jonathan Rodden has observed. The electorate is typically equal parts Democrat and Republican at about 900 people per square mile, according to Mark Muro of Brookings. The exact number varies a bit from place to place; higher in more Republican and lower in more Democratic states. Overall, majorities tend to flip from blue to red roughly where commuter suburbs give way to “exurban” sprawl. That’s where the political boundary of the density divide is drawn.

Higher population density predicts higher Democratic vote share even in small cities in deep red counties in deep red states.

Wilkinson explains this “density divide” by suggesting the process of urbanization sorts people with relatively fixed prior characteristics (ethnicity, ethnocentrism, personality type, educational attainment) into different geographies, and then ghettoization within now more homogeneous places further entrenches those differences. Wilkinson’s perspective is worthy of careful consideration. I’m not writing to dispute or endorse his view, but to offer an alternative explanation that might or might not complement it. (Our tendency to like debates hot—and to attach our egos to this view or that—often leaves us arguing hypotheses as though they are mutually exclusive when, in social affairs, they usually are not.)

A very simple explanation for Wilkinson’s density divide might have to do with people’s intuitions about the pervasiveness of externalities. People who live in dense places correctly perceive that the things their neighbors do blow back upon them. If a neighbor wants to tear-down her single family home and replace it with a midrise apartment building, the typical urbanite doesn’t think “it’s her property, she can do what she wants”. She goes straight to the planning commission to complain about parking and shadows. If someone wants to open up a strip joint, a gun range, even a posh bar in a dense city, urbanites instinctively understand that those choices will have consequences for people other than the entrepreneur and her clients, and demand a regulatory process to balance the external costs (and benefits, at least in theory) against the benefits that business stakeholders would enjoy. This view, that most potential actions have significant repercussions for people who do not voluntarily choose to participate in them, we’ll call externality pessimism.

In more rural areas, libertarian intuitions seem more sensible. For many actions, external costs decline pretty directly with distance. In places sufficiently sprawled that parking is rarely an issue but nearly all interaction is mediated by an automobile journey, a kind of hermetically sealed, climate-controlled pod-world serves as both an insulator and diffuser. The external cost to you if some kind of eyesore becomes your “neighbor” is much less if your neighbor is well down the road and you just have to drive past it than if it is literally adjacent to your home. The dangers you (or at least wussy urbanites like me) might fear from rough clientele at a gun range are attenuated if the clientele doesn’t actually congregate anywhere in particular, because once they step from the parking lot into their vehicles they are everywhere and nowhere at once. Gun ownership itself looks very different from the perspective of someone living in the countryside, for whom policing and eyes-on-the-street mutual supervision cannot be relied upon and an unintended stray bullet is unlikely to hit anything more vulnerable than a tree, than within a city. In an urban context, the external costs of pervasive gun ownership are large. Weapons might fall into the hands of lunatics and criminals, and there are always lunatics and criminals nearby. Incautious shots intended recreationally could harm or kill someone. The petty conflicts that are ubiquitous in urban life, that in the ordinary course of things flare hot and then are quickly forgotten, might needlessly escalate to tragedy, when the satisfying punctuation of a trigger is near at hand. People who live in sprawl or genuinely rural places just don’t bump into one another as much. Their homes are their castles. Driveways and garages, cul-de-sacs and long ribbons of asphalt, serve as moats. A real sphere of privacy exists, where in general the effects, good and bad, of people’s actions are mostly restricted to those within the household. If we accept that, except in extreme circumstances, households are best placed to see to their own interests, then there is not much call or place for external regulation. What people do with and on their own property is, to a good approximation, their own affair, and meddling by outsiders, whether well-intended or corrupt, is likely to do more harm than good. This is externality optimism. If the costs and benefits of people’s choices are fully internalized, they can be left to do whatever they wish.

If we think not so much about the current, Trumpified Republican party (there will be another lens for that!), but the coalition that used to flatter itself as the Party of Reagan, Wilkinson’s density divide becomes largely explainable in terms of where people understandably sit between externality optimism and pessimism. By virtue of direct lived experience, people living at high densities perceive nearly all actions as interactions and seek (though rarely find) a competent and fair regulatory framework to mediate conflicts of interest. People living at low densities experience a world much closer to Econ 101 models: interaction is not the default, and when it is sought, markets can coordinate mutually beneficial outcomes. External regulation is counterproductive.

Many of my readers will, I think, see this as an unduly charitable read of even the pre-Trump Republican Party. Maybe. One of my aims in this exercise is to find bases for some charity between factions, without which I think we have a great deal to fear. Still, let’s go through why I also think the pre-Trump Republican Party was misguided in its externality optimism, despite (I claim) how understandable that optimism may have been from the lived experience of individuals who affiliate with that party.

Once upon a time, a person’s direct lived experience may have served as a sufficient proxy for the forces that affect people’s lives. But that time has long passed. Technology and the scale at which our economy and finance has become organized mean that we all live in a crowded city, whether we notice it or not. Urbanites, in a sense, have an unearned advantage because our direct experience forms intuitions that better match a reality that in fact we all share than those that might be formed in salubrious isolation on a Wyoming ranch. Consider this phenomenal story by Claire Kelloway (ht Matt Stoller) on the tremendous squeeze Big Ag has placed on predominantly Republican farmers.

Market concentration is what I refer to as a power externality. At every step of the way, firms like Monsanto may have gained their market strangleholds via notionally voluntary transactions, all parties to which believed in their interest. But the accumulated effect of those transactions is to increase the bargaining power of just a few firms and undermine the bargaining power of future counterparties, who may not have consented to any of the earlier transactions. For example, much of what is sought by the acquirer in a merger is a better “market position”. That benefit is internal to the transaction: Both the acquirer and the target often understand its economic value, and price negotiations determine how the benefit will be shared. But the benefit of a better market position, which translates to increased monopoly or monopsony power, is created by increases in revenue from future customers or reductions of payments to future vendors, all of whom are not parties to the merger. Antitrust regulation, in this sense, can be understood to be much the same as land use regulation in cities. It exists to balance the interests of transactors who internalize benefits against costs the transaction will impose upon third parties.

Farmers aren’t stupid. They fully understand how and by whom they are being squeezed, and their ex-urban perspective on gun rights notwithstanding, countering the power of Big Ag might be a way for Democrats to make inroads into rural America (as Kelloway argues). This kind of vulnerability helps to explain Republican operatives’ enthusiasm for emphasizing “culture”, which allows them to highlight the small, the kind of controversies rank-and-file voters might experience very locally, and call attention to the absurdity and perniciousness of meddling by governments or self-appointed do-gooders into matters where the rural peace would best be kept by leaving people alone and perhaps building good fences.

Climate change is another domain where I think the externality lens can help us understand how the parties have sorted. Droughts, violent storms, floods, and heat waves do not discriminate between rural people and city dwellers. These catastrophes are indifferent to the culture wars that flamboyantly mark our tribes. But people from lower density environments just intuitively have less reason to believe in weird, counterintuitive consequences of apparently benign actions. Urbanites are constantly arguing over prima facie good things. (Housing for the humans is good but it is bad!) They frequently make and consider counter-intuitive cases with appeals to indirect effects. That’s less common for people who live at lower densities. Republican politicians almost constantly emphasize—because their constituents are receptive to them—notions like “common sense”. A simple version of common sense often works pretty well in low density environments where the (directly perceptible) consequences of most actions are experienced by the people who voluntarily participate in them. People formed by lower densities have stronger priors to overcome before they concede the existence of invisible but urgent consequences to what seem like ordinary, and very valuable, activity. Given the stakes for America’s huge oil and gas industry, (at least) one of America’s two political parties was always liable to be climate reactionary, regardless of the evidence. Of the two parties, perhaps climate skepticism finds a more natural home among Republicans than Democrats not because rank-and-file Republicans are anti-science, but because their reasonably formed priors render the burden of evidence for “spooky action at a distance” higher than for density-addled Democrats.

It is easy for people like me (and most likely for people like you, dear reader, although I hope not all of you) to conclude that it is Republicans who are basically mistaken, afflicted with an externality optimism that may be understandable but is in fact unsupportable in the contemporary world. I think that’s true, but not quite the right lesson to draw. For while externality pessimism is more suited, descriptively, to the modern world than externality optimism, in a prescriptive sense there is perhaps no project more important than restoring or creating contexts within which externality optimism would be correct and adaptive.

Wilkinson writes, “There are no Republican cities.” An obvious retort by Republicans might be, “There are no well-governed cities.” Among larger cities, at least within the United States, I think the retort is broadly correct. You don’t have to endorse the slander against my remarkable native city by this moment’s unfortunate President to concede that the problem of effective governance remains pressing and unsolved in America, at the municipal level as much as any other. America’s most prosperous cities are far from immune. Governance is the world’s most pressing unsolved problem. We need new ideas and “petrie dishes” that enable creative experimentation and hopefully progress. We are, for now, just terrible. The scale and scope of diverse modern polities has outstripped our capacity to govern.

So, arguably one of the best approaches we have for now, is to try to reorganize our affairs so we need governance less. That’s a quite different claim than to argue for “small government” in the world as it is. As poorly governed as our cities currently are, it would be idiocy to argue for less government where in fact the reality of pervasive externalities demands extensive regulation and negotiation. But, to the limited degree that it is possible without making ethically abhorrent or materially intolerable trade-offs, encouraging ways of organizing ourselves that render externality optimism a less deluded intuition would mitigate some of the harms of our incapacity to govern.

The first-best solution, of course, would be to get better at governing, at every scale and level. If the problem of governance were solved, if we knew how to i) coordinate the provision of public goods at any scale, with ii) widespread legitimacy and minimal dispute over the distribution of benefits and obligations iii) across diverse constituencies while iv) upholding liberal values, the possibilities that would be unlocked for human flourishing are unfathomable. And though we may never fully achieve all that, governance is a practice, an institution, a social technology. It is susceptible to improvement. We can make progress.

But in the meantime, altering circumstances where we can, so we require governance less, so that the stakes of political conflict are lower, may be a helpful stopgap. That sounds like, and is, a very Republican kind of insight. But it’s also the basis for a lot of small-is-beautiful intuitions among lefties. And it’s the basis for these words, by John Maynard Keynes in 1933 (ht Tyler Cowen):

To begin with the question of peace. We are pacifist today with so much strength of conviction that, if the econornic internationalist could win this point, he would soon recapture our support. But it does not now seem obvious that a great concentration of national effort on the capture of foreign trade, that the penetration of a country’s economic structure by the resources and the influence of foreign capitalists, and that a close dependence of our own economic life on the fluctuating economic policies of foreign countries are safeguards and assurances of international peace. It is easier, in the light of experience and foresight, to argue quite the contrary. The protection of a country’s existing foreign interests, the capture of new markets, the progress of economic imperialism–these are a scarcely avoidable part of a scheme of things which aims at the maximum of international specialization and at the maximum geographical diffusion of capital wherever its seat of ownership. Advisable domestic policies might often be easier to compass, if the phenomenon known as “the flight of capital” could be ruled out. The divorce between ownership and the real responsibility of management is serious within a country, when, as a result of joint stock enterprise, ownership is broken up among innumerable individuals who buy their interest to-day and sell it to-morrow and lack altogether both knowledge and responsibility towards what they momentarily own. But when the same principle is applied internationally, it is, in times of stress, intolerable—I am irresponsible towards what I own and those who operate what I own are irresponsible towards me. There may be some financial calculation which shows it to be advantageous that my savings should be invested in whatever quarter of the habitable globe shows the greatest marginal efficiency of capital or the highest rate of interest. But experience is accumulating that remoteness between ownership and operation is an evil in the relations among men, likely or certain in the long run to set up strains and enmities which will bring to nought the financial calculation.

I sympathize, therefore, with those who would minimize, rather than with those who would maximize, economic entanglement among nations. Ideas, knowledge, science, hospitality, travel—these are the things which should of their nature be international. But let goods be homespun whenever it is reasonably and conveniently possible, and, above all, let finance be primarily national. Yet, at the same time, those who seek to disembarrass a country of its entanglements should be very slow and wary. It should not be a matter of tearing up roots but of slowly training a plant to grow in a different direction.

Do read the whole thing. It is a remarkable essay.

We are left, I suppose, with externality pessimism of the intellect, externality optimism of the will. Until we get better at governing.

Update History:

  • 27-Apr-2019, 3:40 p.m. EEST: “Technology and the scale at which our economy and finance has become organized meansmean that we all live in a crowded city”; “…the retort is broadly truecorrect.”; “externality pessimism of the mindintellect, externality optimism of the will”; “But oneOne of my aims in this exercise…”; “TheThese catastrophes are indifferent”; “…once they step from the parking lot into their vehiclevehicles…”

Prologue: Lenses

Lately I’m thinking a lot in terms of what I call “lenses”. By lenses, I mean something quite similar to, but a bit mushier than, what Julia Galef and her colleagues call a “double crux“. In a nutshell, a double crux is a narrow point of contention that can be found to account for a broader disagreement. If we are arguing about, say, whether it is wise to permit construction of high-end condominium towers in increasingly unaffordable cities, we might find that we would agree with our opponent if we shared their view on the largely empirical question of “induced demand”. One party opposes permitting the towers, because she believes that building them will draw in new high-income residents from elsewhere, doing little for affordability as new supply is matched to new demand. The other party believes that demand conditions are not so much affected by high-end new construction, so the new units would be taken by current residents, allowing their older units to “filter” as new supply to lower income clienteles. The two agree that if they took the opposite view on this narrow question, they would switch sides on the broader question.

There are a lot of virtues in finding double cruxes, but the one I will emphasize has to do with a kind of intellectual charity. By identifying a narrow, relatively anodyne source of disagreement, the parties take what feels like a value-laden, almost tribal conflict and reframe the disagreement in terms of a judgment call people might understandably disagree about. Instead of the rival parties coming to see themselves as almost alien to one another — one virtuous, the other “bought” in some fashion, one reasonable the other impervious to logic — the two parties reframe one another as rational people with similar values who disagree over an unclear fact about the world.

In practice, in the context of a hot argument, I think it is pretty difficult to find clean double cruxes. Even when one has putatively been agreed, one or both sides is likely to accuse the other of being impervious to the evidence they can marshal on that new, narrower question, and so of being unreasonable or worse after all. For the most part, I think, the humans come to their judgments via a complicated miasma of personal perceptions, interests, and values, second-hand evidence (“scientific” and otherwise), communal identities, and idiosyncratic intuition. Syllogistic argument serves more as a backfilled means of by which we try to summarize and communicate all that than as the source of our views. Before we can actually be persuaded by syllogistic argument (sometimes we can!), we have to find a new equilibrium that reconciles and integrates the new views into the complicated psychosocial ecology that forms us and that we help to form. I don’t think this is necessarily bad. It may often be the case that beliefs and perceptions that emerge from this complicated miasma are more functional, even more accurate, than those we produce trying to be evidence-based and rational. Formal rationality is limited by the scope of the tools we invent as forms of reason and the inputs we consecrate as admissible evidence. That is likely to constitute an almost infinitessimal fraction of the space of potential decision-making processes. Our messy ids were ruthlessly selected by natural and social selection to manage existential threats to our individual and small-group existences. There is no doubt that our “guts” can be very badly misled under circumstances — like the formalized, media-saturated, large-scale societies in which we presently subsist — that diverge from where they evolved. And even when our guts work as advertised, they may produce judgments that might in some narrow sense be functional but that fall ethically outside of contemporary norms (and so are no longer adaptive in contexts where those norms are socially enforced). Using physiognomic markers to discriminate between in- and out-groups may have been a functional strategy among warlike hunter-gatherer tribes, but represents a kind of racism most of us today hope to ensure is not adaptive to pursue in contemporary contexts. Nevertheless, I think we know of ourselves (and, more recently, of the “artificial intelligences” we now produce in our image) that restricting decision-making to what we can elaborate or justify using tools of formal rationality is simply inadequate to the task of producing decisions in real time that are functional in the world. The other stuff is unruly and sometimes awful. But we need it.

“Lenses” are my attempt to adapt the admirable charity of looking for double cruxes to this more naturalistic account of belief formation and decision-making. Rather than a specific conjecture, disagreement about which is sufficient to account for a particular dispute, a “lens” is a broad intuition that I think differs across social and intellectual tribes. Like double cruxes, these intuitions are more anodyne, more obvious to all as judgment calls about which people might understandably differ, than the bitter disputes they serve some role (I claim) in motivating. Rather than explain variation in positions on specific disputes, they help explain how and why we sort ourselves into communities of radically divergent and contending worldviews.

Identifying a lens isn’t the same as saying there is no right or wrong answer. Some broad intuitions about the world may be more factually supportable, and others plainly mistaken. More commonly, some intuitions are right in particular contexts and wrong in others, and much of our tribalism may be explained by extrapolating too universally from “guts” that were correct in some accustomed circumstance, but that fail when extended in time, over geography, or across social groups. Part of the fun of this project is identifying my own intuitions, and imagining how varying those might place me into different groups with different worldviews.

I hope that thinking this way can help facilitate a kind of cross-group empathy. Some of the social fissures and political conflicts we’re currently experiencing derive from differences in material interests or deeply-held values that no amount of empathy can bridge. But most of our conflicts are not that, I think, except indirectly in the sense that those whose material interests are threatened sometimes work to inflame conflicts among people whose interests might otherwise be reconcilable and opposed to their own.

I don’t, by the way, think there is anything original in this, and to the degree there is anything good in it I’m happy to attribute it to Galef et al’s project, which I’ve admired for a while. My “divergences of broad intuition” might be reframed as “different distributions of priors”, and could be identified as plain old double cruxes under a Bayesian rather than syllogistic rationality. “Lenses” are also perhaps related to Arnold Kling’s “axes“.

More than a thousand words have passed and I suspect that you, dear reader, still have no clue what I am talking about. Things may (or may not!) become clearer when I provide some examples of “lenses”. Let’s, um, see! Very soon, I hope.

Update History:

  • 27-Apr-2019, 12:35 p.m. EEST: “It’sIt may often be the case that beliefs and perceptions that emerge from this complicated miasma are more functional…”