Tavis D. Jules
With the advent of the fourth industrial revolution and the intelligent economy, this conceptual chapter explores the evolution of educational governance from one based on governing by numbers and evidence-based governance to one constituted around governance by data or data-based educational governance. With the rise of markets and networks in education, Big Data, machine data, high-dimension data, open data, and dark data have consequences for the governance of national educational systems. In doing so, it draws attention to the rise of the algorithmization and computerization of educational policy-making. The author uses the concept of “blitzscaling”, aided by the conceptual framing of assemblage theory, to suggest that we are witnessing the rise of a fragmented model of educational governance. I call this governance with a “big G” and governance with a “small g.” In short, I suggest that while globalization has led to the deterritorializing of the national state, data educational governance, an assemblage, is bringing about the reterritorialization of things as new material projects are being reconstituted.
INTRODUCTION
Today, data is heralded as the “new oil” of our generation and a vital curator of the so-called “Intelligent Economy” that is premised upon the application of intelligent algorithms, knowledge, innovation, and the internet economy (see Fig. 1). The transition from governments to markets and the evolution of market-based economies to knowledge-based economies implies that the new sources of wealth are intelligence in the form of information housed in clouds, harnessed through data procedures, broken down into uniquely tailored bites, and sold off to the highest bidder. The algorithmization and computerization of educational policy-making have arrived, and it is differentiated by data-driven governance (as opposed to evidence-based policy-making) where data is now multimodal; the source, analysis, output, and evaluator. With the progression of cloud computing, we have witnessed the arrival of the “measurement generation” which counts in zettabytes (one-sextillion bytes), yottabytes (one-septillion bytes), brontobytes (one-octillion bytes), and geopbyte (one-nonillion bytes). In today’s internet-driven industrial revolution, or the so-called Fourth Industrial Revolution, data in any of its 13 forms – Big Data,1 structured, unstructured, and semi-structured data,2 time-stamped data,3 machine data,4 spatiotemporal data,5 open data,6 dark data,7 real-time data,8 genomics data,9 operational data,10 high-dimension data,11 unverified outdated data,12 translytic data,13 (Bridgwater, 2018; Huang & Jin, 2018; Li, Feng, Chin Ooi, Wang, & Zhou, 2011) – is multiplying as it digitally transforms the ways in which we live, work, govern, and educate. This new economy is energized by digital technologies and has at its core intelligent machines and sensors ranging from self-learning algorithms to interconnected devices. As the parameters of innovation evolve, we are witnessing the transition from the technological paradigm of the knowledge-based economy to industry 4.0, which is characterized by the integration of Big Data with the production process (OECD, 2018). With the move from the linear economy to the circular economy, the race to denominate machine learning, a subfield of artificial intelligence, is energized by who can commodify, consume, and customize the most significant amount of data for clients.
With the emergence of this new oil, questions around data privacy have arisen in the wake of several scandals (ranging from social [Facebook], to military [Fitness App Polar], to political [Cambridge Analytica Ltd.]) as to who governs the different aspects of data. Following Cope and Kalantzis (2016), Big Data in education is viewed as
the purposeful or incidental recording of activity and interactions in digitally mediated, network-interconnected learning environments – the volume of which is unprecedented in large part because the data points are smaller, and the recording is continuous. (p. 2, emphasis in original)
Fig. 1. The Intelligent Economy.
While these discussions have now come to the forefront of national and global policy agendas, the impact of data governance on the educational sector has rarely been discussed. When data governance and education are talked about in the same vein, they are always linked to workforce development, educational data mining, and learning analytics (Cope & Kalantzis, 2016; Gagliardi, Parnell, & Carpenter-Hubin, 2018; Williamson, 2017) or the ways in which disruptive innovation has challenged the orthodoxy of the classroom (Christensen, Horn, & Johnson, 2008).
The amount of data collected and mined about why students perform better in some international tests over others means that Comparative and International Education is now big business. For example, in the USA, in 2017, $2.7 billion was invested into ed-tech companies (up from $1.6 billion in 2016) by venture-capital investors, while millions have been spent on implementing technology-based personalized learning by the Silicon Valley Community Foundation, the Gates Foundation, and the Chan-Zuckerberg Initiative (Atlantic, 2018). With the rise of markets and networks in education, Big Data, machine data, high-dimension data, open data, and dark data have consequences for the governance of national educational systems. For the last few decades, “educational governance tools” (Jules, 2012) and “governance mechanisms” (Dale, 1999) (as opposed to markets or hierarchy) have gained legitimacy as they “organize and carry out governing interactions in the face of diversity, complexity, and dynamics” (Kooiman, Bavinck, Chuenpagdee, Mahon, & Pullin, 2008, p. 5).
The role of the state is changing, and in instances where the state has been perceived as failing in its governance responsibility, new actors have entered the conversation, and in education these actors now include public–private partnerships that use “interactive governance” by seeking “to solve societal problems and to create societal opportunities; including the formulation and application of principles guiding those interactions and care for institutions that enable and control them” (Kooiman et al., 2005, p. 17). In other words, we are now in an era in which governing is akin to governance and the “public” is replaced with the “private” as part of hybrid partnership configurations involving state and non-state actors (Ball, 2019; Levi-Faur, 2012; Robertson, Mundy, Verger, & Menashy, 2012). As Jules (2017) suggests, this now implies that the global educational policy environment is gated, regulated, and “over” governed since “newer actors” or “education brokers” are advancing a different set of educational governance mechanisms and several newer modus operandi (or modes, styles, and arrangements) of governance – collaborative governance,14 interactive governance,15 network governance,16 global experiential governance,17 meta-level governance,18 performance-based governance,19 evidence-based governance,20 and data-driven governance – which have the ability to make decisions in real time. In this way, educational governance mechanisms have orthodoxly been viewed as driven by external policy mechanisms of influence; however, the combination of technology and data is revolutionizing educational decision-making processes and educational governance.
Thus, I am suggesting that with the adoption of new technologies, Big Data is reterritorializing educational governance. These new territories are global in scope, scale, and dynamics since today’s national policy environment is multiscalar, multispatial, and multilayered and guided by new state and non-state actors. With the movement to a student-as-customer model in education, educational systems are now capturing consumer behavior and preferences to make analytical predictions about what and how to govern. In what follows, I first highlight an overview of the historical archives and dissemination of data in comparative education by “international knowledge banks” (Jones, 2007). Next, I explore how the transition from data collection to data-based evidence policy-making has respaced national educational governance. In the second half of the paper, I use the concept of “blitzscaling” aided by the conceptual framing of assemblage theory to suggest that we are witnessing the rise of a fragmented model of educational governance. I call this governance with a “big G” and governance with a “small g.” In short, I argue that while globalization has led to the deterritorializing of the national state, data educational governance, an assemblage, is bringing about the reterritorialization of things as new material projects are being reconstituted.
HISTORY OF DATA COLLECTION AND GOVERNANCE IN COMPARATIVE AND INTERNATIONAL EDUCATION
In Comparative and International Education (CIE), Maroy (2009) asserts that “governance models” indicates “theoretical and normative models serving as cognitive and normative references, especially for decision-makers, in defining ‘good ways to steer or govern’ the education system” (p. 76). The collection, commodification, and consumerism of data in CIE is nothing new, and has been done since the World Bank’s first educational loan to Tunisia in 1962, and has been used in large part to drive data-based decision-making and evidence-based policy-making choices ever since. Under these strategic priorities that highlighted evidentiary policy-making, governments lacked the analytical policy capacity to manage the policy process, and these were then outsourced to “international knowledge banks” (Jones, 2007), such as the World Bank and the International Monetary Fund, which later applied loan conditionalities to education. This is described by Jones (2007) as the rise of “educational fundamentalism” that later transformed into “educational multilateralism” (Mundy, 1998) and now exists as mechanisms of coordination under “educational regionalism” (Jules, 2015). With the movement toward the assessment of educational opportunities, outcomes, monitoring, and the production of reports, we have entered an era defined and driven by digital educational products and services.
Data Collection and the International Evaluation of Educational Achievement (IEA)
One of the oldest data collection, commodification, and consumerist entities in CIE is the IEA, which was created in the late 1950s and conducted its first assessment of learning in the early 1960s (Lindblad, Pettersson, & Popkewitz, 2018). Prior to IEA,
comparing education had been undertaken more from out of humanistic ideals, but with the formation of the IEA by scientists with interest in psychometrics and with an outspoken interest in educational outputs, social sciences and behavioral science came to be the ideal on which comparative achievement tests rested. (Lindblad et al., 2018, p. 3)
Today, IEA uses country-based sampling and is one of the largest collectors of educational data in the form of large-scale comparative studies on policy, curricula, and student outcomes. IEA’s large-scale international assessment (ILSA) data and other studies are made accessible to researches through its online “Gateway.” ILSA’s data play a pivotal role in providing economic and social policy guidance to countries (Mølstad & Pettersson, 2018). As Lindblad et al. (2018) further state,
[…] the IEA created something new in the history of comparing education. It focused on educational output that could be represented in numbers: they created hierarchies of students and educational systems as well as nations based on these numbers, and as a result, IEA created specific positivistic reasoning on education. (p. 4)
IEA’s advancement into the collection of educational data began with the First International Mathematics Study (FIMS), conducted in 12 countries in 1964, which featured samples of 13-year-old students and pre-university students and focused on how teaching and learning in mathematics influences societal, scientific, and technological change (IEA, 2019). While FIMS at this point focused solely on mathematics, its overall aim “was to examine the differential output of school systems, using achievement in mathematics as the independent variable” (Robitaille, 1990, p. 396). Data for FIMS led to the Second International Mathematics Study (SIMS), which was conducted in 22 countries in the early 1980s and also assessed students aged 13 and those who were finishing their secondary schooling (Horvath, 1987). As IEA (2019) suggests, the aim was to examine mathematics education in middle schools across three dimensions: curricula, classroom practices, and student achievement in the final year of schooling. In the early 1990s, the Computers in Education Study (COMPED) collected data on the impact of the introduction of computers in participating countries. Around the same time, the 1990–1991 Reading Literacy Study, conducted in 32 countries, began mining data and using it to have states enact evidence-based, data-driven decision-making.
Beginning in 1995, IEA and the United States by way of the National Center for Education Statistics (NCES), began c...