Tag Archives: capitalism

Understanding Class: Reflections on Erik Olin Wright’s Multi-layered Synthesis

Class is a key category and concept in political economy. In saying this, I am not saying that class is the totality or the cause of all inequality, or the base of all economic processes. The days of seeing all history as the history of class struggle are long gone, but the distribution of wealth among different classes of people, and the understanding of class behaviour, mobilisation and power remain of critical importance – both to political economy and to society generally.

Yet despite (or because of) this importance, the concept and analysis of class is hotly contested and muddied by different meanings and analytical frameworks. In this context I recently came across a reference to Erik Olin Wright’s attempt to bring together Marxist, Weberian and Durkheimian definitions and theories of class into a coherent (rather than competing) multi-layered framework. (While Wright is famous, I have not read much of his work – not sure why!) The following is my summary and reflections – informed as always by some concrete campaign problems I am currently thinking about.

Photo of book: Understanding Class, by Erik Olin Wright

Understanding Class

In one of his last books, Understanding Class (2015), American sociologist Erik Olin Wright starts with the broadest of framing of different class traditions:

  • Marxist: where class is a system of exploitation structured around the accumulation of wealth based on the ownership of capital;
  • Weberian: where class is a location within market relations built on the hoarding of economic privileges to the exclusion of others (think: education credentials or licencing requirements for middle-class jobs, union closed shops or the exclusive rights of private property);
  • Durkheimian: where class is a collection of attributes and life conditions which determine people’s position in society (e.g. geographic location, occupation, cultural traits, health, housing conditions – all of which impact on access to income and resources).

My first experience of these differences was a long time ago when a mad activist sought to verify my proletarian credentials by grilling me about my parents’ occupation and where I went to school. I replied that I did not own the means of production and had to sell my labour power. Talking at cross-purposes.

However, Wright brings these broad approaches together by seeing them as simply operating at different levels and asking different questions. This is best explained by a metaphor of a game. Marxism is interested in what game is being played (capitalism or socialism). Weberian class is about the rules of the game (how the system is regulated and maintained – with various regulatory regimes being possible within capitalism). The Durkheimian tradition is about moves within the rules of the game – with individuals or interest groups vying within a given set of rules to get a better slice of the pie.

I have adapted one of Wright’s tables (and associated discussion) to summarise the framework.

LevelGame MetaphorPolitical Focus/IssuesClass Analysis
SystemicWhat game to playRevolutionary v counter-revolutionary politicsMarxist
InstitutionalRules of the gameUnion regulation, employer rights/control, role of the stateWeberian
SituationalMoves in the gameGovernment spending, tax rates, subsidiesDurkheimian

Wright takes his analysis in a strange sociological direction with lots of graphs of class politics and political strategy. However, applying his multi-layer model to local politics, I would suggest a mud map of left-of-centre class politics as follows:

  • the NGO left operates mostly at the level of situational inequality (in part because they are formed and funded around particular issues),
  • left-leaning economists, the “industrial left” and the majority of those identifying as “socialist” operate mostly at the institutional level (eg. social democracy v neoliberalism), while
  • everyone claims to be operating at the systemic level!

Critique

As a tool of class analysis, I have a few issues with Wright’s framework. It implies that Marxism has little role in institutional struggles around tax, distribution and regulatory regimes. Yet an analysis of the processes of accumulation is surely central to such politics.

Further, it reduces the “systemic” choice to whole of society changes (revolution or counter-revolution), rather than seeing capitalism as just one particular set of production relations among many – albeit the dominant set (although as noted in an earlier post, even this hegemony is debatable in a digital age). From this perspective, the system-choice politics at the top level of Wright’s analysis may be less than revolutionary – everything from debates over public ownership/privatisation of utilities (i.e. production by the state or by capitalist businesses), to whether a forest is protected (i.e. kept outside production) or felled as a commodity for a capitalist production process. And, most mundanely, the question of whether to cook dinner at home (non-commodified household production) or to eat out (capitalist production) is a system-choice.

Finally, some things will have some characteristics or elements which could fit in to different categories. For instance, a minor concession reform to extend support to someone with a particular attribute (e.g. low paid, precarious work) may simply be a situational move which does not change the position or power of the recipient. However, it still implies and calls for a particular redistributive role for the state – which is an operation at the institutional level. And as Wright notes, the cumulative impacts of a number of situational moves/policies at some point could amount to or force institutional reform.

The Usefulness of the Analysis

However, despite these concerns, the multi-layered framework in Wright’s Understanding Class is important and useful for a number of reasons. Firstly, it reminds us that while we may be using similar terms, such as “class”, we may mean very different things and have very different understandings of what that means. But more importantly, it highlights the incompleteness of any of the paradigms.

This has very direct implications for policy development and what to do about inequality. To focus simply on systemic issues and institutions which promote or challenge class inequality is to ignore the very different ways those structures and institutions impact on people in different situations. Conversely, to simply focus on the attributes of people or their differing situations risks victim-blaming (it is about their attributes) and not challenging the institutions and structures which drive inequality.

As an example, I am thinking about my current work developing a campaign for more public housing. It seems to me that if we simply argue on the basis that certain groups are excluded from housing, then even if we can identify these groups by particular attributes (e.g. older women, ex-prisoners, unemployed youth), we will not get more public housing unless we also win the institutional campaign around the role of the state in providing housing – and perhaps we won’t win that debate without making inroads into the systemic ideology around the limitations of a capitalist market. That said, if we are too focused on big picture debates about economic systems, people won’t see a relevance to their particular circumstances and we won’t mobilise the support needed to change anything. And of course, most organisations/campaigns simply don’t have the resources to operate at all three levels.

Further, it is not even as simple as which level to focus on. There are also questions about the language and data we use. To take a different example, if we define poverty or inequality as being based on household income (the mainstream definitions) and then to do analysis of the dis/proportion of women in poverty, or on different educational levels or sources of income within the poverty cohort, we are talking (often unintentionally) a Durkheimian language of attributes rather than systems (in the Marxist sense) or the institutional arrangements (in the Weberian sense) which drive that poverty and inequality. To talk about those things, we may need different data (e.g. the wage share of GDP, but much more besides) and a different language.

Conclusion – of sorts

Like Erik Olin Wright, I originally came from a particular camp in the great class debate, but now want to acknowledge the power and potential of different perspectives – even if I don’t quite know how to develop this in practice. How can we organise politically at multiple levels – especially if we don’t have a shared language or understanding? What might class alliances look like if they are splintered by the different (and potentially conflicting) attributes and situations that people bring to confronting the institutions and structures of class (and other inequalities for that matter)? Indeed, according to one reading of his last book (2019), Wright himself came to believe that the working class had become too fractured to play the role Marxism traditionally assigns to it.

Or there is the question I ask myself most often at work: how do we address structural inequality if the language, data and rules of the game (including views of what is possible) are already set to a limited field of situational inequality?

I occasionally see opportunities to use data differently, to ask different questions or challenge orthodoxies or unconscious theoretical propositions, but I don’t have a macro-theory or big answers – particularly to the question of organising diversity.

However, despite my questions and lack of answers, I think Wright’s approach in Understanding Class is interesting and provocative. It will be the background I will bring to reading a forthcoming book edited by Steven Threadgold and Jessica Gerard on Class in Australia, which I have just pre-ordered. I look forward to seeing what frameworks are engaged and at what level we might progress. UPDATE: Post on Class in Australia here.

Understanding the Digital Economy: Two Books, Two Views

This post starts as review of Shoshana Zuboff’s The Age of Surveillance Capitalism, considers a different view from McKenzie Wark, and ends up querying whether this is really capitalism and the political implications of such debates.

Surveillance Capitalism

Zuboff’s monumental 600-page book argues that with digital technology we are in a new phase of capitalism – an age of “surveillance capitalism”. She is clear that although it is a market form which would be unimaginable outside the digital milieu, surveillance capitalism is not reducible to nor determined by digital technology. The reluctant admission of a former Google CEO that search engines retain [personal] information for “quite some time” is an exemplar: it is not the search engine that retains the data, it is those who wrote the programs and command/enable this “surveillance capitalism”.

Zuboff situates surveillance capitalism as historically specific. While the (ch 2) history is fairly light (a zeitgeist change with little reference to power or the interests behind the ideological shifts she describes), the basic argument is that neoliberalism stripped the old corporate forms of inefficient social compromises (like collective bargaining, awards, welfare states) and created a new business focus on “shareholder value” (profit maximisation). This, combined with new individualised consumption wants created the moment for the “Apple ipod hack” (i.e. personalised consumption on demand) and the birth of surveillance capitalism.

In this vein, Zuboff cites Piketty on the rising inequality, but argues that the real clash is between the experience of increasing inequality and the modern individualised consciousness with an expectation of making one’s own destiny – a clash between feudal-level wealth distribution and more egalitarian modern thought. This is evident in the Occupy Movement’s fixation on “the 1%” which by definition was not a revolution for the marginalised and oppressed, but for a supermajority of the 99% of society.

Chapter 3 outlines the key political economy of surveillance capitalism in a shift from what began as an original behavioural value exchange (you give google/algorithm behavioural data and they provide a better, more personalised service) to a surplus extraction. Along with your search/click/usage history, a whole lot of surplus data is collected about you (and combined with data that is scraped from across the web). This data is marketised through algorithms and AI to predict future behaviour to maximise advertising returns (by marketing the product to a relevant consumer at the time they are most likely wanting that product).

Note: Google (which is Zuboff’s main exemplar) is not selling your data, they are selling their prediction of your consumption interests in what is in effect a behavioural futures market. Google’s original genius was the discovery of a way to translate non-market interactions (e.g. information searches) into a raw material (behavioural surplus) for products (behaviour predictions) aimed at their real customers (the advertisers).

Chapter 4 take this basic microeconomic analysis into a macro political economy drawing on Polanyi’s famous analysis of how the market transforms human life into “labour”, nature into “land” or “real estate”, and exchange into “money”. Similarly, under surveillance capitalism human experience is commodified into a market process which goes way beyond the consumer’s own interactions or interests.

This is a really interesting framework to which I will return, and at this point I expected the book to begin detailing how this process took place and to trace the contours of this commodification and the accumulation of wealth. Unfortunately though, it took a very different direction.

Much of the technology story and corporate history is treated as assumed knowledge (Zuboff argues it has been well-described but under-theorised in lots of places). Instead the book seeks to add further layers of analysis building on psychology. It argues that surveillance capital now captures not just actions (searches, clicks, views) but also reactions and emotions. It is then able to not only predict, but to “nudge”/change behaviour – a result that limits free will and removes the “will to will” at an industrial scale.

For me, this is where the problems start for the book. It is rhetoric on rhetoric, and the main evidence is two case studies: a facebook experiment which nudged people to vote in a US election (just to turn-out, it did not track who they voted for), and the Pokemon Go phenomena which moved people around the city colonising both public and private spaces (often at advertisers’ direction) and gathered further locational data. This all seemed at the trivial end of behavioural modification and the “end of free will” argument seemed over-blown. People were still making their own decisions (albeit nudged) to vote or to transverse the city – but doesn’t all advertising try to nudge behaviour?

Further, the argument that there is an ongoing drive for more information accumulation is oft-repeated, but it is never really explained why this data capture needs to be exponentially expanded. Certainly, better data and more accurate predictions give commercial advantage, but is this really a uniquely surveil-or-die imperative? And while obviously digital firms like Google, facebook, Apple and Amazon are huge, there is no real discussion of how this surveillance capitalism mixes with or applies to the rest of the economy. (A point dealt with much better by Wark – see below).

At this point Zuboff’s argument could have focused much more on the political rather than commercial implications, but she does not go there until Ch 13 where again the government use of these technologies is only mentioned in passing, or by reference to China’s social credit regime. This is genuinely scary, but a long way from surveillance capitalism. In fact, I got more political relevance from an excellent report by Monash University academics that looked at the impact of audience segmentation and advertising.

They argues that we are literally getting different information from the same searches depending on our profiles, thus reinforcing our views and limiting any shared understanding. We are not just dealing with different views or values, we are working with completely different facts and we don’t even know what facts others are getting. This contributes to political and social fracturing and is a major challenge for democracy and governance. But Zuboff does not really go there, because as one reviewer points out – she is a professor in Harvard Business School and her interest is business and consumers, not political power.

Anyway Chapter 11 contains a good summary of the argument to date and how they did it – with 16 reasons why they get away with it.[2] Some of this elevates commercial tactics (e.g. law suits, normalising) into a grander conspiracy and it never really rises above innuendo or simply the playbook of (non-digital) big tobacco and big polluters. However, chapter 12 tries to develop a more theoretical argument to buttress the concerns about free will, and goes back to BF Skinner’s radical behaviouralist psychology (Zuboff was a young academic in Skinner’s department). This is the belief that there is no free will or individual agency, just responses to a range of stimulus (with the belief in free will holding science back).

In Skinner’s time they did not have the technology to collate and analyse all the stimuli to which people were responding, so could not predict with certainty, but surveillance capitalism now does have those tools. Data is rendered to the objective observer of Skinner’s dreams – the algorithmic “Big Other”. This is different from Orwell’s Big Brother because Zuboff wants to distinguish between totalitarianism (which seeks to wins souls to a belief system) and surveillance capitalism which does not care about beliefs as long as the data can be rendered so that stimulus can be applied (for market purposes).

This chapter (and ultimately the book) was scary, but left me wanting more. Again, there was the lack of evidence of behavioural change that actually took away free will beyond simply being a more personalised way of affecting what all advertising and cultural “nudging” does. (All culture constrains free choice because it imposes a cost of going outside the normal, or constrains visions of “other” so that one does not know what is unknown – the power of the non-decision that Bachrach and Baratz were noting in the 1970s).

In a lengthy (and at times irrelevant or extravagant review) in The Baffler, Evgeny Morozov goes further, noting that the notion of an appropriation of behavioural surplus from users is misleading. Unlike extraction of money or labour time, the fact that a computer knows I like avocado on toast does not mean that I cease to know it. I have lost nothing in this transaction and continue to enjoy breakfast, so it is not appropriation.

More generally, Morozov argues that Zuboff focuses almost exclusively on the “surveillance” part of surveillance capitalism and left the “capitalism” part of the term un-explored. They trace Zuboff’s earlier work and how it is situated within the Parsonian systems theory and a school of business history which tends toward narrow functionalist explanations (changes arise to address dysfunctionalities).

This analysis is blind to macro-economic power (because the focus is on business and consumers who naturally and unproblematically exist in relation to eachother). Indeed, the focus of surplus extraction shifts from the production process (Marx’s extraction of surplus labour) to an extraction from consumers – without a real consideration of the labour processes that lay behind this commodification process.

Finally, when Morozov eventually focuses on the Zuboff’s book (in section IX of the review!), he offers three ways of interpreting Surveillance Capitalism. The first is that it is largely descriptive with no great claim to revealing a new system dynamic. In this case, it is not clear why all the bother just to reveal a few harms done by Google and facebook, if there are no broader implications.

The second thesis is closer to how I read Zuboff, which is that surveillance capitalism is unequivocally worse than other digital regimes and it is becoming the new hegemonic form of capitalism. However, for both Morozov and me, the evidence for this is lacking (though the review goes into much more detail on how this is lacking).

The third thesis is really a circular argument that surveillance capitalism is driven to behave in an accumulative way. However, the particulars and the results are more drawn from the traditional logic of profit-maximising capitalism than from unique data extraction processes.

In short, Zuboff’s book fails because it is narrowly focused on market-exchanges rather than a broader understanding of capitalism as a production and accumulation process. (On top of that, I will add that it was really long-winded with constantly repeating rhetorical flourishes in place of actual evidence).

Beyond Capitalism?

However, in light of these issues, the next book I read was much more interesting – particularly if viewed in the light of Zuboff’s arguments up to chapter 3. McKenzie Wark’s Capital is Dead: Is this Something Worse? suggests that the sorts of digital market processes that Zuboff describes[3] actually constitute new relations of production and accumulation which are not capitalist.

Unlike Zuboff, Wark comes out of a Marxist tradition and does focus on production relations (not market exchanges). Wark argues that the wealth derived in these digital businesses does not come from the appropriation of surplus labour from a waged proletariat or from the ownership of capital (as per capitalist production relations), but from:

  • The surplus information/data supplied for free by consumers, not as a commodity in a capitalist value-exchange; and
  • The intellectual/technical output of a “hacker workforce” which provides the knowledge for the algorithms and data management, but whose creative work is developed at any time and place and not confined to an industrial/office process.

Wark recognises that the digital machines are produced in standard capitalist industries, and there are a range of other capitalist industries in play, but the point is that the surplus of industrial owners is increasingly wrested by the “vector class” who own the means of information management. We (the left) are so schooled in a Marxian historical story that capitalism would be replaced by socialism (either evolutionary or revolutionary) that we are not seeing new production relations which are not capitalist.

The moment of the proletarian revolution was lost (or perverted) early last century, but it is worth remembering that it was not the peasants that overthrew feudalism. Rather it was capitalists that supplanted feudal lords (over time, sometimes with force/law, sometimes with greater efficiency). Similarly, according to Wark, the new vector class who control information is supplanting those who own capital as the hegemonic class. Capitialism and capitalist production relations continue to exist (just as non-capitalist forms of production continued under capitalism), but they are no longer the dominant part of the economy.

Unfortunately, IMHO, at this point in the book Wark goes off into cultural arguments rather than further considering the production relations, but the argument around production relations took me back to my PhD reading and the “Rethinking Marxism” school. Over 30 years ago they theorised capitalism as a specific production relation (rather than an over-arching “mode of production”) and as only one among a number of existing production and class processes.

I was surprised/frustrated that Wark did not draw on that thinking at all, but when I went to the Rethinking Marxism journal there were (unsurprisingly) a range of articles on how to theorise digital technology and production. After an introductory note about “magic” and technological fetishism, these articles ranged from seeing simply “variants of capitalism as usual” to considerations of “more-than-capitalism”: digital feudalism (extraction of rent in the form of data), knowledge creation and non-class processes of accumulation, and a potential to strengthen the commons/community.

Unfortunately, most of the articles in that issue are not online, so that rich reading must wait for another day when I can get to a university library.

POLICY IMPLICATIONS

Ultimately though, I am less interested in the labels (capitalism/not-capitalism) than in the analysis and its implications. Despite the gaps in both Zuboff and Wark’s books, both argue that our thinking and legislation has not caught up with the new digital world. This is important because it seems to me that unless we can understand the logic of accumulation in a digital economy, our political responses will be limited, or flawed.

For instance, traditional consumer issues like privacy and consumer data rights, while important, are based on liberal capitalist concerns which struggle for relevance in a digital age. Similarly, the horrors of digital surveillance in predictive policing, court sentencing and welfare compliance that Virginia Eubanks details in Automating Inequality are a serious problem, but addressing this won’t disturb the new logic of accumulation.

If we are interested in political economy from a perspective of class, inequality and a fair sharing of wealth, and if the digital economy is indeed based on new forms of power and wealth creation, then we need to go beyond the legalistic issues above and build new social regulation and redistribution mechanisms.

A starting point could be the combination of new and old tax measures of the type proposed by Thomas Piketty (Capital and Ideology, Ch 17), or new measures such as the proposal from Yanis Varoufakis to make big tech companies pay into a public wealth fund with a capital return to every citizen (in recognition of the value of the public’s data that underscores the tech giants’ profits). These proposals provide a positive agenda, but it is noteworthy that even Varoufakis’ radical proposal is based on the logic of capitalism (the rights to a return on capital, which in this case is data). Accordingly, the transformative nature of these proposals may also be limited.

Perhaps that is ok, but it probably depends on whether we understand the globalised digital economy as a continuation of the economy we have known or something fundamentally different. On this, I suspect there is more exploration to be done.

[2]              Maybe just read Ch11 – it would be much quicker than wading through the whole book!

[3]              This is my shorthand. Wark only refers to Zuboff in passing and along with others who have theorised new forms of digital/platform capitalism.