Yanis Varoufarskis’ book, Technofeudalism: What Killed Capitalism, draws heavily on two books that I began this blog series discussing: Shoshana Zuboff’s The Age of Surveillance Capitalism and McKenzie Wark’s Capital is Dead: Is This Something Worse?. Given this background, I was more than interested to hear Varoufarkis talk at Adelaide Writers’ Week, and to read the book.

In my reading, Varoufarkis takes from Zuboff an understanding of the economic importance of digital data and the ability of corporations to nudge (or more) consumer behaviour to reshape the market and profit. From Wark, he takes the idea that this is more than a new phase of capitalism and is a different form of production.

My goal here is not to properly review the Varoufarkis’ book. Rather, after a brief explanation of the central thesis (skip to here you are familiar with it) I want to add some thoughts from my own theoretical journey in political economy which I think add credence, or at least possibility, to the concept of technofeudalism.

The Technofeudalism Thesis

To me, Varoufarkis takes the best of Zuboff and Wark, but goes beyond this and builds a more convincing argument by showing how control of data enables its owners to accumulate wealth (and political power) – essentially via rent drawn from the production and exchange which takes place in privately defined and owned spaces in the online cloud.

Just as feudal lords enclosed the commons and extracted a portion of the surplus for the right to live and farm on their estates, so too those who own the cloud fiefdoms (Amazon, Google, Facebook, etc) collect rent from all those whose business models require the public access to those virtual fiefdoms and the information generated within them. Capitalists seeking a profit are trapped in these new tech fiefdoms (because of the high price of exiting in loss of exposure, customers, and profit) in the same way that feudal serfs were trapped on the landlords’ estate.

But others have questioned whether this is really happening at scale and whether it is really not capitalism?

Varoufarkis has a variety of answers:

• the fact that goods and services are still being produced in capitalist firms does not mean that capitalism is dominant, any more than the rise of a capitalist class in earlier centuries meant that agriculture and land rents disappeared.
• the data that drives the system is formed from the capture of the thoughts, desires and needs of “cloud serfs” (all of us – knowingly or unknowingly) who provide the data for free.
• This data capture can be expanded exponentially with little new investment, so financial return is not a return on capital, but rather a rental charge for others accessing this information and operating in this (privatised) space.

It is a dominance of rent over profit as the primary centre of accumulation that signals that capital has been supplanted.

Thoughts

1. Dominant Sectors

Weirdly, thinking about technofeudalism replacing capitalism took me back to historical debates I studied as an undergraduate many years ago. The question was how to characterise the Australian economy in the late nineteenth century. From memory, the traditional view of a predominantly pastoral economy (“riding on the sheep’s back”) was challenged by NG Butlin and others who, armed with Kenyesian economics and new national accounting tools, argued that manufacturing contributed more to economic production. However, those arguing in a Marxist tradition insisted on the need to look at the centres of accumulation and capture of state power as the defining dynamic, rather than the share of the economy. In this light, pastoralism and pastoralists remained hegemonic long after the rise of manufacturing.

I have no idea where these historical debates are at now, but the point here is that how we characterise the economy is theory-dependant, and that simple shares of GDP or workforce are not necessarily conclusive.

In this context, when the world’s richest people are those who own the largest slices of cloud capital, it is not unrealistic to question whether the cloud (and the rent from it) is the new centre and the dominant system, even while traditional profit-making industry is more widespread.

2. Cloud serfs and the definition of production and consumption

Varoufarkis’ technofeudal economy is inhabited by “cloud prols” and “cloud serfs”. The former simply appear to be super-exploited workers in an algorithm-dominated, but nonetheless capitalist production process (hence the proletariat label). However, it is the cloud serfs who take us beyond capitalism. The data collection by search engines and digital platforms is a direct appropriation of productive resources – not a market sale of labour power or other productive inputs which characterise the capitalist production process.

But are people using digital platforms (consumers of those digital services) really producing wealth or enabling the capture of surplus to drive a new economic system? It defies a most fundamental distinction in economics between producers and consumers. However, as I argued in my PhD, this distinction was always artificial – a product of a particular neoclassical theory which conflated work, production, the market and the economy, while households and non-market spaces become consumers.

This distinction is hardwired into our definition of the economy and the national accounts which describe it. Yet around half the goods and services produced in Australia are produced outside the market and market production itself consumes labour power which is produced in the household. In this circular flow between households and market production, the split between production and consumption is arbitrary (or at least theory-dependent).

Marxism draws the production/consumption line differently by positing much activity in the market economy as social reproduction or distribution/consumption of surplus, rather than production. However, it is feminist economics which mounts the strongest challenge to the neat distinction between consumers and producers – both because of the non-market production in the household, but also because much of women’s emotional labour is embodied and not alienable as labour power sold in a market.

In short, we should not simply posit payment and the market as the arbiter of what constitutes production and economic processes. We should be able to analyse the appropriation of personal information and data as a systemic part of the accumulation process on its own terms, not bound by the conceptions of theories based on different economic relations. In this context, the term “cloud serf” may be clunky, but it is consistent with the analysis of the basis of accumulation as rent rather than profit.

3. The mode of production and the metanarrative

Much of the debate around the nature of the new digital economy, whether it is surveillance capitalism, cloud capitalism or technofeudalism is based in narratives around modes of production – the description of the economy as a whole and the metanarrative of the transition from feudalism to capitalism to socialism (maybe, eventually!). But as JK Gibson-Graham point outs, this metanarrative is problematic. The whole idea of “the Capitalist system” leaves no outside, no vision of the other in a perspective which limits political and transformative action.

The alternative, within a Marxism perspective, is a more micro focus on class processes – the different ways in which goods and services are produced and how surplus labour is appropriated. These processes are not economy-wide: there is production and surplus appropriation within households (also, in Resnick and Wolf’s analysis, feudal – bound by ties of family, loyalty and tradition rather than market exchanges), in owner-operated enterprises where there is no direct surplus appropriation, and in government and not-for-profit enterprises where any surplus goes to a common good rather than private profit).

In this analysis, any given person, household, community or economy is characterised by the particular intersection of a range of these class (appropriation) processes as well as non-class processes of power and privilege.

The point here is that, even if one was not convinced by Varoufarkis’ argument that technofeudalism is a new and dominant mode of production (as per point 1 above), it is at least fairly obviously a different set of class processes (by virtue of its non-market appropriation of raw materials and its potential expansion based on data rent rather than capital investment). Free of the question of whether it is the dominant/defining mode of production, there is more scope to analyse the technofeudalist processes on their own terms – and to develop an opposition to them specifically, without it being predicated on the overthrow of society as a whole.

I have vacillated for more than 20 years on the usefulness of the analysis of a “capitalist system” verse a focus on specific production relations and appropriation processes – but either way, it seems to me the technofeudalism argument is challenging, important, and a little petrifying.