Calculated publics? Hazards of meritocracy in promising efficiency and rationalism

As search engines and other ranking and sorting software devices promise efficient and rational outcomes, many researchers started to question the calculation or math behind them, including the research interests on “governing algorithms” or Gillespie’s concept on “calculated publics“. However, I believe the answer to that question does not have to be loaded with computer science vocabularies because at the core of the question is the social construct of meritocracy and its hazards for political experiments broadly defined. The best stories that I can think of to illustrate this point are the two recent stories on China’s adoption of the US models of national accounts for economy (e.g. GDP indicators) and global accounts of scholar output (e.g. SCI). Indeed, although China is proud to be “different” from the U.S. with their Chinese characteristics , Chinese use (or misuse) of the GDP and SCI has been religiously performative, even more so than their U.S. counterparts.

Recent researchers’ interest in “governing algorithms” or Gillespie’s concept on “calculated publics” is warranted to provide some critical views of the information systems around users. These information systems promise to deliver “relevant” outcomes efficiently, usually based on user-generated data (e.g. user reviews) and/or user activity (users’ clicks or purchases). I agree that researchers should help users to examine these systems more critically and reflexively. However, I think the concept of “calculability” for governance is way before the Internet and computers. The issue of “calculability” in human societies have been associated with modernity: rationality and bureaucracy in classical sociology.

So if one wants to question the issue of calculability in modern life, particularly in the 20th Century, one cannot miss the indicators of GDP, invented and hailed by the U.S. Department of Commerce as “one of the Great Inventions of the 20th Century“. Indeed it is quoted to be a fundamental information instruments to account for national economic activities, preventing a nation from “economic dark ages”:

Information is fundamental to understanding all human endeavor. The national income accounts, and the data they use and produce, are our core economic information. While they can—and with adequate human, financial, and organizational resources, will—be continually improved; without them we would be in economic dark ages.

Thus, if we take the concept of “governing algorithms” seriously, then it matters also to revisit the GDP indicators, including how these numbers are collected, calculated, circulated and evaluated. One major illustrative case will be the U.S. and Chinese economies. Despite their differences between the so-called Washington Consensus and Beijing Consensus for economic development, both use GDP “algorithm”, at least nominally, to govern their national economy.

In particular, the Chinese GDP numbers are increasingly important not only for the local politics but also for the international politics. The recent news on China’s GDP mis-report or mis-spoke by Chinese officials/media has attracted attention from western and Chinese observers alike. Also, a recent “province-sized hole in GDP figures” news also proves the issue of calculability for governance is well and alive even in Chinese context, despite all claims on “Chinese characteristics”.

At the very least, both China and the U.S. agrees on using the GDP as one (if not “the”) important indicator to run the biggest two (national) economies of the world, (although sometimes Chinese may resort to other more “realistic numbers” so as to avoid the official “performative” acts, as shown in the so-called Keqiang index).

So we can say that the caculability for governance has to be “performative”. The concept of “performativity” may be helpful in situating these numbers and algorithms as speech acts of economic governance. It is legitimate for researchers to question various issues of such caculability-based governance, including its lack of accountability for emotional work or eco-system sustainability, and thus the demand for better alternative “algorithms for human well-being”. Indeed, states in this sense are machines that execute economic algorithms by collecting all kinds of economic data. Whether a state is functioning or not depends on its capacity to run such economic “algorithms” by technocratic experts, as explained by Timothy Mitchell on Egypt.

A similar observation can be made on the academic performance on using the SCI indicators for “impact” evaluation. Again, China beats the U.S. by practising this caculability game (or executing scholar-output-evaluation algorithm) more religiously than the US. The inventor of the SCI, Dr. Eugene Garfield, has visited China many times and was surprised by the way the SCI outcome is (mis)used by Chinese technocrats. Many Chinese scientists have criticized the Chinese misuse of the SCI on various grounds, and some even call it “Stupid Chinese Idea” (SCI) and “Stupid Chinese Idea Ends its Naturally Continuable Evolution” (SCIENCE) as the “end of science”. (For more, read a panel discussion on Reflection on Chinese academic evaluation system from SCI, lead by scientists such as Guo-Jie Li). In China, it has been reported that the scientific publication system is so intertwined with caculability to the point of market economic calculatibilty (i.e. pay to play) produce an industry of its own.

The Chinese use (or perfection) of the GDP and SCI “algorithms” to grow its economic and scholarly output provides an interesting data point for Internet researchers to reflect on how new the concepts of “governing algorithms” or “calculated publics” truly are. Personally I see more historical continuities than departures from the pre-existing modern caculability machinery that has been running/governing our lives. If it is not new, then the questions may well go back to the old concerns on the role of (and most likely the rule by) the experts and technocrats who run these systems. The ultimate question can be boiled down to the modern question of meritocracy.

Can a machine-like system work better to produce a fairer, more efficient social system?

In this sense, China is the natural country that struggles with such a modern question, much earlier than the West, as carefully examined by historian Alexander Woodside in “Lost Modernities: China, Vietnam, Korea, and the Hazards of World History“. In this book, Woodside notes that the Western scholars have only recently recognized that meritocracy, in capitalism and bureaucracy alike, may be one of the major destablising hazard factors itself and thus urges researchers to learn from the Chinese history in maintaining such a meritocracy. In the end, the research question becomes one of the classic question on elites and meritocracy, something similar to the recent critique of the US and global failure in achieving (instead of just preaching) a true meritocratic society in the book of “Twilight of the Elites: America After Meritocracy“.

Still, some technical and substantial departures do exist from the indicators of GDP and SCI to the online algorithms of ranking and social relevance engineering. Internet researchers may and should continue to question the algorithm-based calculability and the social costs/risks behind it. However, I do believe that by revisiting the idea of meritocracy in the longer historical context of human modernization projects, researchers can gain more with some historical insights and more grounded evidence-based departures beyond just using computer-science terminologies such as algorithms or operating systems, including the concepts of “the relevance of algorithms” and “social operating systems

Comments (choose your preferred platforms)

Loading Facebook Comments ...