Digifesto

Category: sociology

Omi and Winant on economic theories of race

Speaking of economics and race, Chapter 2 of Omi and Winant (2014), titled “Class”, is about economic theories of race. These are my notes on it

Throughout this chapter, Omi and Winant seem preoccupied with whether and to what extent economic theories of race fall on the left, center, or right within the political spectrum. This is despite their admission that there is no absolute connection between the variety of theories and political orientation, only general tendencies. One presumes when reading it that they are allowing the reader to find themselves within that political alignment and filter their analysis accordingly. I will as much as possible leave out these cues, because my intention in writing these blog posts is to encourage the reader to make an independent, informed judgment based on the complexity the theories reveal, as opposed to just finding ideological cannon fodder. I claim this idealistic stance as my privilege as an obscure blogger with no real intention of ever being read.

N.B. Sep. 17 2020 – Ironically, I see now that these blog posts get a steady stream of traffic. For my formally published work on this topic, please see  “Racial categories in machine learning”, coauthored with Bruce Haynes.

Omi and Winant devote this chapter to theories of race that attempt to more or less reduce the phenomenon of race to economic phenomena. They outline three varieties of class paradigms for race:

  • Market relations theories. These tend to presuppose some kind theory of market efficiency as an ideal.
  • Stratification theories. These are vaguely Weberian, based on classes as ‘systems of distribution’.
  • Product/labor based theories. These are Marxist theories about conflicts over social relations of production.

For market relations theories, markets are efficient, racial discrimination and inequality isn’t, and so the theory’s explicandum is what market problems are leading to the continuation of racial inequalities and discrimination. There are a few theories on the table:

  • Irrational prejudice. This theory says that people are racially prejudiced for some stubborn reason and so “limited and judicious state interventionism” is on the table. This was the theory of Chicago economist Gary Becker, who is not to be confused with the Chicago sociologist Howard Becker, whose intellectual contributions were totally different. Racial prejudice unnecessarily drives up labor costs and so eventually the smart money will become unprejudiced.
  • Monopolistic practices. The idea here is that society is structured in the interest of whites, who monopolize certain institutions and can collect rents from their control of resources. Jobs, union membership, favorably located housing, etc. are all tied up in this concept of race. Extra-market activity like violence is used to maintain these monopolies. This theory, Omi and Winant point out, is sympatico with white privilege theories, as well as nation-based analyses of race (cf. colonialism).
  • Disruptive state practices. This view sees class/race inequality as the result of state action of some kind. There’s a laissez-faire critique which argues that minimum wage and other labor laws, as well as affirmative action, entrench race and prevent the market from evening things out. Doing so would benefit both capital owners and people of color according to this theory. There’s a parallel neo-Marxist theory that says something similar, interestingly enough.

It must be noted that in the history of the United States, especially before the Civil Rights era, there absolutely was race-based state intervention on a massive scale and this was absolutely part of the social construction of race. So there hasn’t been a lot of time to test out the theory that market equilibrium without racialized state policies results in racial equality.

Omi and Winant begin to explicate their critique of “colorblind” theories in this chapter. They characterize “colorblind” theories as individualistic in principle, and opposed to the idea of “equality of result.” This is the familiar disparate treatment vs. disparate impact dichotomy from the interpretation of nondiscrimination law. I’m now concerned that this, which appears to be the crux of the problem of addressing contests over racial equality between the center and the left, will not be resolved even after O&W’s explication of it.

Stratification theory is about the distribution of resources, though understood in a broader sense than in a narrow market-based theory. Resources include social network ties, elite recruitment, and social mobility. This is the kind of theory of race an symbolic interactionist sociologist of class can get behind. Or a political scientist’s: the relationship between the elites and the masses, as well as the dynamics of authority systems, are all part of this theory, according to Omi and Winant. One gets the sense that of the class based theories, this nuanced and nonreductivist one is favored by the authors … except for the fascinating critique that these theories will position race vs. class as two dimensions of inequality, reifying them in their analysis, whereas “In experiential terms, of course, inequality is not differentiated by race or class.”

The phenomenon that there is a measurable difference in “life chances” between races in the United States is explored by two theorists to which O&W give ample credit: William J Wilson and Douglas Massey.

Wilson’s major work in 1978, The Declining Significance of Race, tells a long story of race after the Civil War and urbanization that sounds basically correct to me. It culminates with the observation that there are now elite and middle-class black people in the United States due to the uneven topology of reforms but that ‘the massive black “underclass” was relegated to permanent marginality’. He argued that race was no longer a significant linkage between these two classes, though Omi and Winant criticize this view, arguing that there is fragility to the middle-class status for blacks because of public sector job losses. His view that class divides have superseded racial divides is his most controversial claim and so perhaps what he is known best for. He advocated for a transracial alliance within the Democratic party to contest the ‘racial reaction’ to Civil Rights, which at this point was well underway with Nixon’s “southern strategy”. The political cleavages along lines of partisan racial alliance are familiar to us in the United States today. Perhaps little has changed.
He called for state policies to counteract class cleavages, such as day care services to low-income single mothers. These calls “went nowhere” because Democrats were unwilling to face Republican arguments against “giveaways” to “welfare queens”. Despite this, Omi and Winant believe that Wilson’s views converge with neoconservative views because he doesn’t favor public sector jobs as a solution to racial inequality; more recently, he’s become a “culture of poverty” theorist (because globalization reduces the need for black labor in the U.S.) and believes in race neutral policies to overcome urban poverty. The relationship between poverty and race is incidental to Wilson, which I suppose makes him ‘colorblind” in O&W’s analysis.

Massey’s work, which is also significantly reviewed in this chapter, deals with immigration and Latin@s. There’s a lot there, so I’ll cut to the critique of his recent book, Categorically Unequal (2008), in which Massey unites his theories of anti-black and anti-brown racism into a comprehensive theory of racial stratification based on ingrained, intrinsic, biological processes of prejudice. Naturally, to Omi and Winant, the view that there’s something biological going on is “problematic”. They (being quite mainstream, really) see this as tied to the implicit bias literature but think that there’s a big difference from implicit bias due to socialization vs. over permanent hindbrain perversity. This is apparently taken up again in their Chapter 4.

Omi and Winant’s final comment is that these stratification theories deny agency and can’t explain how “egalitarian or social justice-oriented transformations could ever occur, in the past, present, or future.” Which is, I suppose, bleak to the anti-racist activists Omi and Winant are implicitly aligned with. Which does raise the possibility that what O&W are really up to in advocating a hard line on the looser social construction of race is to keep the hope of possibility of egalitarian transformation alive. It had not occurred to me until just now that their sensitivity to the idea that implicit bias may be socially trained vs. being a more basic and inescapable part of psychology, a sensitivity which is mirrored elsewhere in society, is due to this concern for the possibility and hope for equality.

The last set of economic theories considered in this chapter are class-conflict theories, which are rooted in a Marxist conception of history as reducible to labor-production relations and therefore class conflict. There are two different kinds of Marxist theory of race. There are labor market segmentation theories, led by Michael Reich, a labor economist at Berkeley. According to this research, when the working class unifies across racial lines, it increases its bargaining power and so can get better wages in its negotiations with capital. So the capitalist in this theory may want to encourage racial political divisions even if they harbor no racial prejudices themselves. “Workers of the world unite!” is the message of these theories. An alternative view is split labor market theory, which argues that under economic pressure the white working class would rather throw other races under the bus than compete with them economically. Political mobilization for a racially homogenous, higher paid working class is then contested by both capitalists and lower paid minority workers.

Reflections

Omi and Winant respect the contributions of these theories but think that trying to reduce race to economic relations ultimately fails. This is especially true for the market theorists, who always wind up introducing race as an non-economic, exogenous variable to avoid inequalities in the market.

The stratification theories are perhaps more realistic and complex.

I’m most surprised at how the class-conflict based theories are reflected in what for me are the major lenses into the zeitgeist of contemporary U.S. politics. This may be because I’m very disproportionately surrounded by Marxist-influenced intellectuals. But it is hard to miss the narrative that the white working class has rejected the alliance between neoliberal capital and low-wage immigrant and minority labor. Indeed, it is arguably this latter alliance that Nancy Fraser has called neoliberalism. This conflict accords with the split labor market theory. Fraser and other hopeful socialist types argue that a triumph over identity differences is necessary to realize racial conflicts in the working class play into the hands of capitalists, not white workers. It is very odd that this ideological question is not more settled empirically. It may be that the whole framing is perniciously oversimplified, and that really you have to talk about things in a more nuanced way to get real headway.

Unless of course there isn’t any such real hope. This was an interesting part of the stratification theory: the explanation that included an absence of agency. I used to study lots and lots of philosophy, and in philosophy it’s a permissible form of argument to say, “This line of reasoning, if followed to its conclusion, leads to an appalling and untenable conclusion, one that could never be philosophically satisfying. For that reason, we reject it and consider a premise to be false.” In other words, in philosophy you are allowed to be motivated by the fact that a philosophical stance is life negating or self-defeating in some way. I wonder if that is true of sociology of race. I also wonder whether bleak conclusions are necessary even if you deny the agency of racial minorities in the United States to liberate themselves on their own steam. Now there’s globalization, and earlier patterns of race may well be altered by forces outside of it. This is another theme in contemporary political discourse.

Once again Omi and Winant have raised the specter of “colorblind” policies without directly critiquing them. The question seems to boil down to whether or not the mechanisms that reproduce racial inequality can be mitigated better by removing those mechanisms that are explicitly racial or not. If part of the mechanism is irrational prejudice due to some hindbrain tick, then there may be grounds for a systematic correction of that tick. But that would require a scientific conclusion about the psychology of race that identifies a systematic error. If the error is rather interpreting an empirical inequality due to racialized policies as an essentialized difference, then that can be partially corrected by reducing the empirical inequality in fact.

It is in fact because I’m interested in what kinds of algorithms would be beneficial interventions in the process of racial formation that I’m reading Omi and Winant so closely in the first place.

bodies and liberal publics in the 20th century and today

I finally figured something out, philosophically, that has escaped me for a long time. I feel a little ashamed that it’s taken me so long to get there, since it’s something I’ve been told in one way or another many times before.

Here is the set up: liberalism is justified by universal equivalence between people. This is based in the Enlightenment idea that all people have something in common that makes them part of the same moral order. Recognizing this commonality is an accomplishment of reason and education. Whether this shows up in Habermasian discourse ethics, according to which people may not reason about politics from their personal individual situation, or in the Rawlsian ‘veil of ignorance’, in which moral precepts are intuitively defended under the presumption that one does not know who or where one will be, liberal ideals always require that people leave something out, something that is particular to them. What gets left out is people’s bodies–meaning both their physical characteristics and more broadly their place in lived history. Liberalism was in many ways a challenge to a moral order explicitly based on the body, one that took ancestry and heredity very seriously. So much a part of aristocratic regime was about birthright and, literally, “good breeding”. The bourgeois class, relatively self-made, used liberalism to level the moral playing field with the aristocrats.

The Enlightenment was followed by a period of severe theological and scientific racism that was obsessed with establishing differences between people based on their bodies. Institutions that were internally based on liberalism could then subjugate others, by creating an Other that was outside the moral order. Equivalently, sexism too.
Social Darwinism was a threat to liberalism because it threatened to bring back a much older notion of aristocracy. In WWII, the Nazis rallied behind such an ideology and were defeated in the West by a liberal alliance, which then established the liberal international order.

I’ve got to leave out the Cold War and Communism here for a minute, sorry.

Late modern challenges to the liberal ethos gained prominence in activist circles and the American academy during and following the Civil Rights Movement. These were and continue to be challenges because they were trying to bring bodies back into the conversation. The problem is that a rules-based order that is premised on the erasure of differences in bodies is going to be unable to deal with the political tensions that precisely do come from those bodily differences. Because the moral order of the rules was blind to those differences, the rules did not govern them. For many people, that’s an inadequate circumstance.

So here’s where things get murky for me. In recent years, you have had a tension between the liberal center and the progressive left. The progressive left reasserts the political importance of the body (“Black Lives Matter”), and assertions of liberal commonality (“All Lives Matter”) are first “pushed” to the right, but then bump into white supremacy, which is also a reassertion of the political importance of the body, on the far right. It’s worth mention Piketty, here, I think, because to some extent that also exposed how under liberal regimes the body has secretly been the organizing principle of wealth through the inheritance of private property.

So what has been undone is the sense, necessary for liberalism, that there is something that everybody has in common which is the basis for moral order. Now everybody is talking about their bodily differences.

That is on the one hand good because people do have bodily differences and those differences are definitely important. But it is bad because if everybody is questioning the moral order it’s hard to say that there really is one. We have today, I submit, a political nihilism crisis due to our inability to philosophically imagine a moral order that accounts for bodily difference.

This is about the Internet too!

Under liberalism, you had an idea that a public was a place people could come to agree on the rules. Some people thought that the Internet would become a gigantic public where everybody could get together and discuss the rules. Instead what happened was that the Internet became a place where everybody could discuss each other’s bodies. People with similar bodies could form counterpublics and realize their shared interests as body-classes. (This piece by David Weinberger critiquing the idea of an ‘echo chamber’ is inspiring.) Within these body-based counterpublics each form their own internal moral order whose purpose is to mobilize their body-interests against other kinds of bodies. I’m talking about both black lives matter and white supremacists here, radical feminists and MRA’s. They are all buffeting liberalism with their body interests.

I can’t say whether this is “good” or “bad” because the moral order is in flux. There is apparently no such thing as neutrality in a world of pervasive body agonism. That may be its finest criticism: body agonism is politically unstable. Body agonism leads to body anarchy.

I’ll conclude with two points. The first is that the Enlightenment view of people having something in common (their personhood, their rationality, etc.) which put them in the same moral order was an intellectual and institutional accomplishment. People do not naturally get outside themselves and put themselves in other people’s shoes; they have to be educated to do it. Perhaps there is a kernal of truth here about what moral education is that transcends liberal education. We have to ask whether today’s body agonism is an enlightened state relative to moral liberalism because it acknowledges a previously hidden descriptive reality of body difference and is no longer so naive, or if body agonism is a kind of ethical regress because it undoes moral education, reducing us to a more selfish state of nature, of body conflict, albeit in a world full of institutions based on something else entirely.

The second point is that there is an alternative to liberal order which appears to be alive and well in many places. This is an order that is not based on individual attitudes for legitimacy, but rather is more about the endurance of institutions for their own sake. I’m referring of course to authoritarianism. Without the pretense of individual equality, authoritarian regimes can focus on maintaining power on their own terms. Authoritarian regimes do not need to govern through moral order. U.S. foreign policy used to be based on the idea that such amoral governance would be shunned. But if body agonism has replaced the U.S. international moral order, we no longer have an ideology to export or enforce abroad.

Notes on Omi and Winant, 2014, “Ethnicity”

I’m continuing to read Omi and Winant’s Racial Formation in the United States (2014). These are my notes on Chapter 1, “Ethnicity”.

There’s a long period during which the primary theory of race in the United States is a theological and/or “scientific” racism that maintains that different races are biologically different subspecies of humanity because some of them are the cursed descendants of some tribe mentioned in the Old Testament somewhere. In the 1800’s, there was a lot of pseudoscience involving skull measurements trying to back up a biblical literalism that rationalized, e.g., slavery. It was terrible.

Darwinism and improved statistical methods started changing all that, though these theological/”scientific” ideas about race were prominent in the United States until World War II. What took them out of the mainstream was the fact that the Nazis used biological racism to rationalize their evilness, and the U.S. fought them in a war. Jewish intellectuals in the United States in particular (and by now there were a lot of them) forcefully advocated for a different understanding of race based on ethnicity. This theory was dominant as a replacement for theories of scientific racism between WWII and the mid-60’s, when it lost its proponents on the left and morphed into a conservative ideology.

To understand why this happened, it’s important to point out how demographics were changing in the U.S. in the 20th century. The dominant group in the United States in the 1800’s were White Anglo-Saxon Protestants, or WASPs. Around 1870-1920, the U.S. started to get a lot more immigrants from Southern and Eastern Europe, as well as Ireland. These often economic refugees, though there were also people escaping religious persecution (Jews). Generally speaking these immigrants were not super welcome in the United States, but they came in at what may be thought of as a good time, as there was a lot of economic growth and opportunity for upward mobility in the coming century.

Partly because of this new wave of immigration, there was a lot of interest in different ethnic groups and whether or not they would assimilate in with the mainstream Anglo culture. American pragmatism, of the William James and Jown Dewey type, was an influential philosophical position in this whole scene. The early ethnicity theorists, who were part of the Chicago school of sociology that was pioneering grounded, qualitative sociological methods, were all pragmatists. Robert Park is a big figure here. All these guys apparently ripped off W.E.B. Du Bois, who was trained by William James and didn’t get enough credit because he was black.

Based on the observation of these European immigrants, the ethnicity theorists came to the conclusion that if you lower the structural barriers to participation in the economy, “ethnics” will assimilate to the mainstream culture (melt into the “melting pot”) and everything is fine. You can even tolerate some minor ethnic differences, resulting in the Italian-Americans, the Irish-Americans, and… the African-American. But that was a bigger leap for people.

What happened, as I’ve mentioned, is that scientific racism was discredited in the U.S. partly because it had to fight the Nazis and had so many Jewish intellectuals, who had been on the wrong end of scientific racism in Europe and who in the U.S. were eager to become “ethnics”. These became, in essence, the first “racial liberals”. At the time there was also a lot of displacement of African Americans who were migrating around the U.S. in search of economic opportunities. So in the post-war period ethnicity theorists optimistically proposed that race problems could be solved by treating all minority groups as if they were Southern and Eastern European immigrant groups. Reduce enough barriers and they would assimilate and/or exist in a comfortable equitable pluralism, they thought.

The radicalism of the Civil Rights movement broke the spell here, as racial minorities began to demand not just the kinds of liberties that European ethnics had taken advantage of, but also other changes to institutional racism and corrections to other racial injustices. The injustices persisted in part because racial differences are embodied differently than ethnic differences. This is an academic way of saying that the fact that (for example) black people often look different from white people matters for how society treats them. So treating race as a matter of voluntary cultural affiliation misses the point.

So ethnicity theory, which had been critical for dismantling scientific racism and opening the door for new policies on race, was ultimately rejected by the left. It was picked up by neoconservatives through their policies of “colorblindness”, which Omi and Winant describe in detail in the latter parts of their book.

There is a lot more detail in the chapter, which I found quite enlightening.

My main takeaways:

  • In today’s pitched media battles between “Enlightenment classical liberalism” and “postmodern identity politics”, we totally forget that a lot of American policy is based on American pragmatism, which is definitely neither an Enlightenment position nor postmodern. Everybody should shut up and read The Metaphysical Club.
  • There has been a social center, with views that are seen as center-left or center-right depending on the political winds, since WWII. The adoption of ethnicity theory into the center was a significant culture accomplishment with a specific history, however ultimately disappointing its legacy has been for anti-racist activists. Any resurgence of scientific racism is a definite backslide.
  • Omi and Winant are convincing about the limits of ethnicity theory in terms of: its dependence on economic “engines of mobility” that allow minorities to take part in economic growth, its failure to recognize the corporeal and ocular aspects of race, and its assumption that assimilation is going to be as appealing to minorities as it is to the white majority.
  • Their arguments about colorblind racism, which are at the end of their book, are going to be doing a lot of work and the value of the new edition of their book, for me at least, really depends on the strength of that theory.

Notes on Racial Formation by Omi and Winant, 2014, Introduction

Beginning to read Omi and Winant, Racial Formation in the United States, Third Edition, 2014. These are notes on the introduction, which outlines the trajectory of their book. This introduction is available on Google Books.

Omi and Winant are sociologists of race and their aim is to provide a coherent theory of race and racism, particularly as a United States phenomenon, and then to tell a history of race in the United States. One of their contentions is that race is a social construct and therefore varies over time. This means, in principle, that racial categories are actionable, and much of their analysis is about how anti-racist and racial reaction movements have transformed the politics and construction of race over the course of U.S. history. On the other hand, much of their work points to the persistence of racial categories despite the categorical changes.

N.B. Sep. 17 2020 – These informal notes were part of the process of writing “Racial categories in machine learning”, with Bruce Haynes.

Since the Third Edition, in 2014, comes twenty years after the Second Edition, much of the new material in the book addresses specifically what they call colorblind racial hegemony. This is a response to the commentary and question around the significance of Barack Obama’s presidency for race in America. It is interesting reading this in 2018, as in just a few brief years it seems like things have changed significantly. It’s a nice test, then to ask to what extent their theory explains what happened next.

Here is, broadly speaking, what is going on in their book based on the introduction.

First, they discuss prior theories of race they can find in earlier scholarship. They acknowledge that these are interesting lenses but believe they are ultimately reductionist. They will advance their own theory of racial formation in contrast with these. In the background of this section but dismissed outright is the “scientific” racism and religious theories of race that were prevalent before World War II and were used to legitimize what Omi and Winant call racial domination (this has specific meaning for them). Alternative theories of race that Omi and Winant appear to see as constructive contributions to racial theory include:

  • Race as ethnicity. As an alternative to scientific racism, post WWII thinkers advanced the idea of racial categories as reducing to ethnic categories, which were more granular social units based on shared and to some extent voluntary culture. This conception of race could be used for conflicting political agendas, including both pluralism and assimilation.
  • Race as class. The theory attempted to us economic theories–including both Marxist and market-based analysis–to explain race. Omi and Winant think this–especially the Marxist theory–was a productive lens but ultimate a reductive one. Race cannot be subsumed to class.
  • Race as nationality. Race has been used as the basis for national projects, and is tied up with the idea of “peoplehood”. In colonial projects especially, race and nationality are used both to motivate subjugation of a foreign people, and is also used in resistance movements to resist conquest.

It is interesting that these theories of race are ambiguous in their political import. Omi and Winant do a good job of showing how multi-dimensional race really is. Ultimately they reject all these theories and propose their own, racial formation theory. I have not read their chapter on it yet, so all I know is that: (a) they don’t shy away from the elephant in the room, which is that there is a distinctively ‘ocular’ component to race–people look different from each other in ways that are hereditary and have been used for political purposes, (b) they maintain that despite this biological aspect of race, the social phenomenon of race is a social construct and primarily one of political projects and interpretations, and (c) race is formed by a combination of action both at the representational level (depicting people in one way or another) and at the institutional level, with the latter determining real resource allocation and the former providing a rationalization for it.

Complete grokking of the racial formation picture is difficult, perhaps. This may be why instead of having a mainstream understanding of racial formation theory, we get reductive and ideological concepts of race active in politics. The latter part of Omi and Winant’s book is their historical account of the “trajectory” of racial politics in the United States, which they see in terms of a pendulum between anti-racist action (with feminist, etc., allies) and “racial reaction”–right-wing movements that subvert the ideas used by the anti-racists and spin them around into a backlash.

Omi and Winant describe three stages of racial politics in United States history:

  • Racial domination. Slavery and Jim Crow before WWII, based on religious and (now discredited, pseudo-)scientific theories of racial difference.
  • Racial hegemony. (Nod to Gramsci) Post-WWII race relations as theories of race-as-ethnicity open up egalitarian ideals. Opens way for Civil Rights movement.
  • Colorblind racism. A phase where the official ideology denies the significance of race in society while institutions continue to reinforce racial differences in a pernicious way. Necessarily tied up with neoliberalism, in Omi and Winant’s view.

The question of why colorblind racism is a form of racism is a subtle one. Omi and Winant do address this question head on, and I am in particular looking forward to their articulation of the point. Their analysis was done during the Obama presidency, which did seem to move the needle on race in a way that we are still seeing the repercussions of today. I’m interested in comparing their analysis with that of Fraser and Gilman. There seem to be some productive alignments and tensions there.

Inequality perceived through implicit factor analysis and its implications for emergent social forms

Vox published an interview with Keith Payne, author of The Broken Ladder.

My understanding is that the thesis of the book is that income inequality has a measurable effect on public health, especially certain kinds of chronic illnesses. The proposed mechanism for this effect is the psychological state of those perceiving themselves to be relatively worse off. This is a hardwired mechanism, it would seem, and one that is being turned on more and more by socioeconomic conditions today.

I’m happy to take this argument for granted until I hear otherwise. I’m interested in (and am jotting notes down here, not having read the book) the physics of this mechanism. It’s part of a larger puzzle about social forms, emergent social properties, and factor analysis that I’ve written about it some other posts.

Here’s the idea: income inequality is a very specific kind of social metric and not one that is easy to directly perceive. Measuring it from tax records, which short be straightforward, is fraught with technicalities. Therefore, it is highly implausible that direct perception of this metric is what causes the psychological impact of inequality.

Therefore, there must be one or more mediating factors between income inequality as an economic fact and psychological inequality as a mental phenomenon. Let’s suppose–because it’s actually what we should see as a ‘null hypothesis’–that there are many, many factors linking these phenomena. Some may be common causes of income inequality and psychological inequality, such as entrenched forms of social inequality that prevent equal access to resources and are internalized somehow. Others may be direct perception of the impact of inequality, such as seeing other people flying in higher class seats, or (ahem) hearing other people talk about flying at all. And yet we seem comfortable deriving from this very complex mess a generalized sense of inequality and its impact, and now that’s one of the most pressing political topics today.

I want to argue that when a person perceives inequality in a general way, they are in effect performing a kind of factor analysis on their perceptions of other people. When we compare ourselves with others, we can do so on a large number of dimensions. Cognitively, we can’t grok all of it–we have to reduce the feature space, and so we come to understand the world through a few blunt indicators that combine many other correlated data points into one.

These blunt categories can suggest that there is structure in the world that isn’t really there, but rather is an artifact of constraints on human perception and cognition. In other words, downward causation would happen in part through a dimensionality reduction of social perception.

On the other hand, if those constraints are regular enough, they may in turn impose a kind of structure on the social world (upward causation). If downward causation and upward causation reinforced each other, then that would create some stable social conditions. But there’s also no guarantee that stable social perceptions en masse track the real conditions. There may be systematic biases.

I’m not sure where this line of inquiry goes, to be honest. It needs more work.

General intelligence, social privilege, and causal inference from factor analysis

I came upon this excellent essay by Cosma Shalizi about how factor analysis has been spuriously used to support the scientific theory of General Intelligence (i.e., IQ). Shalizi, if you don’t know, is one of the best statisticians around. He writes really well and isn’t afraid to point out major blunders in things. He’s one of my favorite academics, and I don’t think I’m alone in this assessment.

First, a motive: Shalizi writes this essay because he thinks the scientific theory of General Intelligence, or a g factor that is some real property of the mind, is wrong. This theory is famous because (a) a lot of people DO believe in IQ as a real feature of the mind, and (b) a significant percentage of these people believe that IQ is hereditary and correlated with race, and (c) the ideas in (b) are used to justify pernicious and unjust social policy. Shalizi, being a principled statistician, appears to take scientific objection to (a) independently of his objection to (c), and argues persuasively that we can reject (a). How?

Shalizi’s point is that the general intelligence factor g is a latent variable that was supposedly discovered using a factor analysis of several different intelligence tests that were supposed to be independent of each other. You can take the data from these data sets and do a dimensionality reduction (that’s what factor analysis is) and get something that looks like a single factor, just as you can take a set of cars and do a dimensionality reduction and get something that looks like a single factor, “size”. The problem is that “intelligence”, just like “size”, can also be a combination of many other factors that are only indirectly associated with each other (height, length, mass, mass of specific components independent of each other, etc.). Once you have many different independent factors combining into one single reduced “dimension” of analysis, you no longer have a coherent causal story of how your general latent variable caused the phenomenon. You have, effectively, correlation without demonstrated causation and, moreover, the correlation is a construct of your data analysis method, and so isn’t really even telling you what correlations normally tell you.

To put it another way: the fact that some people seem to be generally smarter than other people can be due to thousands of independent factors that happen to combine when people apply themselves to different kinds of tasks. If some people were NOT seeming generally smarter than others, that would allow you to reject the hypothesis that there was general intelligence. But the mere presence of the aggregate phenomenon does not prove the existence of a real latent variable. In fact, Shalizi goes on to say, when you do the right kinds of tests to see if there really is a latent factor of ‘general intelligence’, you find that there isn’t any. And so it’s just the persistent and possibly motivated interpretation of the observational data that allows the stubborn myth of general intelligence to continue.

Are you following so far? If you are, it’s likely because you were already skeptical of IQ and its racial correlates to begin with. Now I’m going to switch it up though…

It is fairly common for educated people in the United States (for example) to talk about “privilege” of social groups. White privilege, male privilege–don’t tell me you haven’t at least heard of this stuff before; it is literally everywhere on the center-left news. Privilege here is considered to be a general factor that adheres in certain social groups. It is reinforced by all manner of social conditioning, especially through implicit bias in individual decision-making. This bias is so powerful it extends not to just cases of direct discrimination but also in cases where discrimination happens in a mediated way, for example through technical design. The evidence for these kinds of social privileging effects is obvious: we see inequality everywhere, and we can who is more powerful and benefited by the status quo and who isn’t.

You see where this is going now. I have the momentum. I can’t stop. Here it goes: Maybe this whole story about social privilege is as spuriously supported as the story about general intelligence? What if both narratives were over-interpretations of data that serve a political purpose, but which are not in fact based on sound causal inference techniques?

How could this be? Well, we might gather a lot of data about people: wealth, status, neighborhood, lifespan, etc. And then we could run a dimensionality reduction/factor analysis and get a significant factor that we could name “privilege” or “power”. Potentially that’s a single, real, latent variable. But also potentially it’s hundreds of independent factors spuriously combined into one. It would probably, if I had to bet on it, wind up looking a lot like the factor for “general intelligence”, which plays into the whole controversy about whether and how privilege and intelligence get confused. You must have heard the debates about, say, representation in the technical (or other high-status, high-paying) work force? One side says the smart people get hired; the other side say it’s the privileged (white male) people that get hired. Some jerk suggests that maybe the white males are smarter, and he gets fired. It’s a mess.

I’m offering you a pill right now. It’s not the red pill. It’s not the blue pill. It’s some other colored pill. Green?

There is no such thing as either general intelligence or group based social privilege. Each of these are the results of sloppy data compression over thousands of factors with a loose and subtle correlational structure. The reason why patterns of social behavior that we see are so robust against interventions is that each intervention can work against only one or two of these thousands of factors at a time. Discovering the real causal structure here is hard partly because the effect sizes are very small. Anybody with a simple explanation, especially a politically convenient explanation, is lying to you but also probably lying to themselves. We live in a complex world that resists our understanding and our actions to change it, though it can be better understood and changed through sound statistics. Most people aren’t bothering to do this, and that’s why the world is so dumb right now.

that time they buried Talcott Parsons

Continuing with what seems like a never-ending side project to get a handle on computational social science methods, I’m doing a literature review on ‘big data’ sociological methods papers. Recent reading has led to two striking revelations.

The first is that Tufekci’s 2014 critique of Big Data methodologies is the best thing on the subject I’ve ever read. What it does is very clearly and precisely lay out the methodological pitfalls of sourcing the data from social media platforms: use of a platform as a model organism; selecting on a dependent variable; not taking into account exogenous, ecological, or field factors; and so on. I suspect this is old news to people who have more rigorously surveyed the literature on this in the past. But I’ve been exposed to and distracted by literature that seems aimed mainly to discredit social scientists who want to work with this data, rather than helpfully engaging them on the promises and limitations of their methods.

The second striking revelation is that for the second time in my literature survey, I’ve found a reference to that time when the field of cultural sociology decided they’d had enough of Talcott Parsons. From (Bail, 2014):

The capacity to capture all – or nearly all – relevant text on a given topic opens exciting new lines of meso- and macro-level inquiry into what environments (Bail forthcoming). Ecological or functionalist interpretations of culture have been unpopular with cultural sociologists for some time – most likely because the subfield defined itself as an alternative to the general theory proposed by Talcott Parsons (Alexander 2006). Yet many cultural sociologists also draw inspiration from Mary Douglas (e.g., Alexander 2006; Lamont 1992; Zelizer 1985), who – like Swidler – insists upon the need for our subfield to engage broader levels of analysis. “For sociology to accept that no functionalist arguments work,” writes Douglas (1986, p. 43), “is like cutting off one’s nose to spite one’s face.” To be fair, cultural sociologists have recently made several programmatic statements about the need to engage functional or ecological theories of culture. Abbott (1995), for example, explains the formation of boundaries between professional fields as the result of an evolutionary process. Similarly, Lieberson (2000), presents an ecological model of fashion trends in child-naming practices. In a review essay, Kaufman (2004) describes such ecological approaches to cultural sociology as one of the three most promising directions for the future of the subfield.

I’m not sure what’s going on with all these references to Talcott Parsons. I gather that at one time he was a giant in sociology, but that then a generation of sociologists tried to bury him. Then the next generation of sociologists reinvented structural functionalism with new language–“ecological approaches”, “field theory”?

One wonder what Talcott Parsons did or didn’t do to inspire such a rebellion.

References

Bail, Christopher A. “The cultural environment: measuring culture with big data.” Theory and Society 43.3-4 (2014): 465-482.

Tufekci, Zeynep. “Big Questions for Social Media Big Data: Representativeness, Validity and Other Methodological Pitfalls.” ICWSM 14 (2014): 505-514.

Values, norms, and beliefs: units of analysis in research on culture

Much of the contemporary critical discussion about technology in society and ethical design hinges on the term “values”. Privacy is one such value, according to Mulligan, Koopman, and Doty (2016), drawing on Westin and Post. Contextual Integrity (Nissenbaum, 2009) argues that privacy is a function of norms, and that norms get their legitimacy from, among other sources, societal values. The Data and Society Research Institute lists “values” as one of the cross-cutting themes of its research. Richmond Wong (2017) has been working on eliciting values reflections as a tool in privacy by design. And so on.

As much as ‘values’ get emphasis in this literary corner, I have been unsatisfied with how these literatures represent values as either sociological or philosophical phenomena. How are values distributed in society? Are they stable under different methods of measurement? Do they really have ethical entailments, or are they really just a kind of emotive expression?

For only distantly related reasons, I’ve been looking into the literature on quantitative measurement of culture. I’m doing a bit of a literature review and need your recommendations! But an early hit is Marsden and Swingle’s is a “Conceptualizing and measuring culture in surveys: Values, strategies, and symbols” (1994), which is a straightforward social science methods piece apparently written before either rejections of positivism or Internet-based research became so destructively fashionable.

A useful passage comes early:

To frame our discussion of the content of the culture module, we have drawn on distinctions made in Peterson’s (1979: 137-138) review of cultural research in sociology. Peterson observes that sociological work published in the late 1940s and 1950s treated values – conceptualizations of desirable end-states – and the behavioral norms they specify as the principal explanatory elements of culture. Talcott Parsons (19.51) figured prominently in this school of thought, and more recent survey studies of culture and cultural change in both the United States (Rokeach, 1973) and Europe (Inglehart, 1977) continue the Parsonsian tradition of examining values as a core concept.

This was a surprise! Talcott Parsons is not a name you hear every day in the world of sociology of technology. That’s odd, because as far as I can tell he’s one of these robust and straightforwardly scientific sociologists. The main complaint against him, if I’ve heard any, is that he’s dry. I’ve never heard, despite his being tied to structural functionalism, that his ideas have been substantively empirically refuted (unlike Durkheim, say).

So the mystery is…whatever happened to the legacy of Talcott Parsons? And how is it represented, if at all, in contemporary sociological research today?

One reason why we don’t hear much about Parsons may be because the sociological community moved from measuring “values” to measuring “beliefs”. Marsden and Swingle go on:

Cultural sociologists writing since the late 1970s however, have accented other elements of culture. These include, especially, beliefs and expressive symbols. Peterson’s (1979: 138) usage of “beliefs” refers to “existential statements about how the world operates that often serve to justify value and norms”. As such, they are less to be understood as desirable end-states in and of themselves, but instead as habits or styles of thought that people draw upon, especially in unstructured situations (Swidler, 1986).

Intuitively, this makes sense. When we look at the contemporary seemingly mortal combat of partisan rhetoric and tribalist propaganda, a lot of what we encounter are beliefs and differences in beliefs. As suggested in this text, beliefs justify values and norms, meaning that even values (which you might have thought are the source of all justification) get their meaning from a kind of world-view, rather than being held in a simple way.

That makes a lot of sense. There’s often a lot more commonality in values than in ways those values should be interpreted or applied. Everybody cares about fairness, for example. What people disagree about, often vehemently, is what is fair, and that’s because (I’ll argue here) people have widely varying beliefs about the world and what’s important.

To put it another way, the Humean model where we have beliefs and values separately and then combine the two in an instrumental calculus is wrong, and we’ve known it’s wrong since the 70’s. Instead, we have complexes of normatively thick beliefs that reinforce each other into a worldview. When we we’re asked about our values, we are abstracting in a derivative way from this complex of frames, rather than getting at a more core feature of personality or culture.

A great book on this topic is Hilary Putnam’s The collapse of the fact/value dichotomy (2002), just for example. It would be nice if more of this metaethical theory and sociology of values surfaced in the values in design literature, despite it’s being distinctly off-trend.

References

Marsden, Peter V., and Joseph F. Swingle. “Conceptualizing and measuring culture in surveys: Values, strategies, and symbols.” Poetics 22.4 (1994): 269-289.

Mulligan, Deirdre K., Colin Koopman, and Nick Doty. “Privacy is an essentially contested concept: a multi-dimensional analytic for mapping privacy.” Phil. Trans. R. Soc. A 374.2083 (2016): 20160118.

Nissenbaum, Helen. Privacy in context: Technology, policy, and the integrity of social life. Stanford University Press, 2009.

Putnam, Hilary. The collapse of the fact/value dichotomy and other essays. Harvard University Press, 2002.

Wong, Richmond Y., et al. “Eliciting Values Reflections by Engaging Privacy Futures Using Design Workbooks.” (2017).