Keywords

Introduction: Crises, Money, and the State

The last decades have been characterized as a time of global crises. There was the Global Financial Crisis (GFC) 2008, and there is at present the Covid-19 pandemic and, in the offing for quite a while, the climate crisis.

The GFC was a crisis emerging from internal contradictions of neoliberalism (which is considered here as a political translation of neoclassical economics), while Covid-19 is generally seen as an external shock, though, obviously through urbanization and globalization, the virus found it easy to spread. The climate crisis, announced already in the 1970s (Club of Rome), increasingly enters the headlines.

All three crises have in common that they force one to rethink the understanding of money and the state. Especially, after decades of austerity politics (accompanied by the mantra “there is no money”), the experience that, in response to the GFC, it was possible to flood the global economy with large sums of money created a widespread cognitive dissonance (“where does all this money suddenly come from?”). Moreover, while the role of the state had been denigrated for a long time as inefficient, the US government (the Treasury and the Federal Reserve Board, FED) had to bail out the banks (though not the people) in order to stabilize the system (Adam Tooze, 2018).

Similar things have happened during the Covid-19 crisis. Governments began pumping large sums into the economy in ways inconsistent with the neoclassical “no-money” mantra. The measures were regarded as temporarily suspending the policies of “solid finance” (keeping government expenditure within the limits of tax income and bond sales). It has already been announced by most governments that all debts will have to be paid back once the Covid-19 pandemic is over. The coming climate crisis, as already understood, will again require governments to find huge sums of money in order to restructure the economy. The climate crisis combines internal and external contradictions and imposes objectives which are unlikely to be met by leaving investment decisions to profitability considerations alone.

These three crises require us to reconfigure the usual understanding of money and the state. Modern Money Theory (MMT) claims to have a different and empirically more grounded concept of money and the state, which allows for identifying a policy space which is far wider than neoclassical economics allows for, and which could accommodate the pressing policy agendas related to these crises.

What has all this to do with distance education (DE)? A short summary of the history of DE from a costing perspective forms the basis for the argument that much of the understanding of the costs and economics of DE has been shaped by (neoclassical) microeconomics with a focus on driving down unit costs and, still consistent with neoclassical policy recommendations, devolving the costs to the learners.

MMT suggests that the microeconomic focus is too narrow, and that embedding the costs and economics of DE and online learning in a macroeconomic context would lead to a more realistic analysis and, consequently, better policy responses. In particular, it advances the notion that devolving cost to the learners, in MMT terms “users of the currency,” is misguided.

A Brief History of DE from a Costing Perspective

The founding of the Open University of the United Kingdom (OUUK) in 1969 will be taken as a point of departure for the purposes of this discussion, although DE is much older than that. DE at the time was intended to widen access to education, much in line with the Universal Declaration of Human Rights (1948), which included a right to education. The Organization of Economic Cooperation and Development (OECD) considered education as an important investment to sustain economic and technological superiority. It was observed that the socialist rival already successfully tapped the cognitive resources of the working class as well as those of women. The Sputnik shock of 1957, when the Soviet Union launched the world’s first artificial satellite, was taken in the West as an alert to reform education in order to widen participation beyond its traditional audience (Papadopoulos, 1994). For the developing world, Coombs called for reforming education to make it more efficient by introducing more division of labor (Coombs, 1985). The health services had successfully started on this path, and the author suggested that education should follow.

In this summarized walk through the history of DE, three phases will be distinguished: Traditional DE, online DE, and ODDE (open, distance, and digital education). The headings are not meant to signify the substitution of one phase by the next but rather are meant to emphasize newly emerging aspects.

Traditional Distance Education

Coombs’ propositions, cited above, were echoed in Peters’ theorization of DE as the “most industrial form of education” (Peters, 1983). Peters was clear that to use technology and division of labor renders education more cost-efficient and, therefore, more accessible.

Traditional DE will be used here as backdrop against which later changes in the cost structure of DE are made visible. The reference frame for analyzing the costs of DE is the cost of traditional campus-based education.

With its lecture halls, libraries, sport facilities, student accommodation, catering, and similar amenities, conventional campus-based higher education (HE) is clearly quite costly, such that it can be taken for granted that the total costs of campus-based education are higher than the total cost of DE (for a given number of students). More interesting is that, even if you strip down total costs to those elements directly impinging on teaching and learning, the costs of DE still can be expected to undercut those of conventional education (Fig. 1).

The obvious structural reasons for the cost efficiency of DE can be identified by revisiting the basic cost equations:

  1. (i)

    TC(N) = F + V*N and

  2. (ii)

    AC(N) = F/N + V

(TC stands for Total cost, F for Fixed cost, V*N for Variable cost, and N for the Number of students. AC for Average cost per student. For N increasing, AC(N) approaches V, the variable cost per student or unit cost.)

Hence, as in many industrial processes, cost efficiency is achieved by a combination of capital for labor substitution and labor for labor substitution (meaning a substitution of more costly by less costly labor) to capture scale economies (SE). How is it possible for DE to achieve this?

DE shifts costs to course development (a fixed cost). The development of DE courses often involves not only the subject matter expert, but also instructional designers and media experts (Koumi, 2006). Course development includes the tests/exam questions as well as evaluation rubrics. Even if, in order to ensure the quality of the course material, course development costs in DE may be higher than in CE (conventional education), total costs of DE rise slower than total costs of CE, such that, for a sufficiently high number of students (N), it holds that TCDE(N) < TCCE(N) (Fig. 1, upper part). Recall, that V (the variable cost per student) determines the inclination of the graph of the (linear) total cost equations. Hence, in order to be cost efficient, it is important for DE that VDE < VCE. To keep V low, DE reduces contact with the teacher and/or “unbundles” the teacher role into various functions, which then can be remunerated differently. Teaching in traditional DE was identified with course development (Mills, 2003, p. 104), which is a fixed cost (contributing to F) and not part of V. Contributing to V is course presentation, as well as student support and the marking of assessments. Hence, it is possible to keep V low by using tutors or adjunct lecturers as compared to professors of subject matter experts. Not only are they remunerated at lower rates, but also often they are on shorter-term contracts and paid per task. In Fig. 1 (lower part), one can observe that the average costs asymptotically fall toward V and that, if VDE < VCE for N sufficiently large, ACDE < ACCE.

Fig. 1
figure 1

Total and average cost equations

Full size image

Two further insights present themselves: first, that AC cannot fall below V (even if you multiply student intake for the course); second, scale economies (SE) do erode, and eventually quite quickly. Planners should not be misguided by the promise of scale economies. Daniel is right that DE can widen access while, at the same time, bring down costs, thus “breaking the iron triangle of cost, access and quality” (Daniel et al., 2009). However, the Daniel argument comes with a sleight of hand: While average costs fall, total costs rise and possibly beyond what the institution can afford (Avoiding the income trap, Butcher, 2004, January, p. 20 ff).

While there are structural reasons for expecting DE to be more cost efficient in terms of cost per student, the same level of cost advantage does often not carry through to comparisons in the cost per graduate (Rumble, 2014). This is due to the often very high dropout rates in DE:

$$ \, \textrm{Cost}\ \textrm{per}\ \textrm{graduate}=\frac{\textrm{cost}\ \textrm{per}\ \textrm{student}}{100\%-\textrm{wastage}} $$

Hence, the cost advantage of DE shrinks due to the often-higher dropout (wastage) rates in DE. The cost efficiency of DE (cost per student) is often so much lower (i.e., “better”) that it can accommodate a considerable level of dropout, before its cost effectiveness falls below that of its conventional competitor (Raza, 2008, pp. 497–498).

While the higher dropout rate in DE could be brushed aside by pointing at the different characteristics of the DE audience, another criticism of DE weighed heavier: the lack of interaction between the student and teacher in DE. Educators quite often see personal interaction as closely linked to the quality and effectiveness of teaching and learning. However, somewhat surprisingly, distance educators also claim interaction for DE. For example, Holmberg argued that, by building some instructional design features into the course material (such as in-text questions and in-text activities) and adopting a more empathetic conversational style in course presentation, the course material itself could be designed as guided didactic conversation, which can trigger an inner monologue which he referred to as “simulated communication” (Holmberg, 1983, 2008). Moore (arguing along similar lines) distinguished three main forms of interaction in DE (Moore, 1989): student-content interaction (SCI), student-teacher interaction (STI), and student-student interaction (SSI). Since not all distance educators found Holmberg’s guided didactic conversation a convincing substitute for student-teacher interaction and interaction among students (Rumble, 2001, p. 3), traditional DE usually included opportunities for face-to-face meetings in evening classes or weekend seminars. Paradoxically, students typically wanted it to be available more than they actually made use of it. It certainly was a factor to drive up the cost of DE.

Online Distance Education

The development of information and communication technologies (ICT) brought new affordances for DE impinging on all three forms of interaction. Hülsmann (building on Rumble 2004) referred to features following from the information-processing aspects of ICT as type-i affordances and to features following from the communication-enabling aspects of ICT as type-c affordances (Hülsmann, 2014, June, 24, p. 244). Type-i affordances include automated responses or simulations, including the learning objects (LO) and, more recently, features based on artificial intelligence (AI) and machine learning, which are applied in learning analytics (LA). All these features can be used to enhance SCI, far beyond the early design features of in-text questions and in-text activities. While much of this (e.g., LA) allowed developing a more precise learner profile, enabling institutions to better support students, it also allowed the close surveillance of students and staff with concomitant ethical issues (Slade & Prinsloo, 2013).

Type-c affordances for the first time in the history of education allowed responsive interaction at a distance, both between teacher and students (STI), and among students (SSI), inaugurating a new paradigm of teaching at a distance (Garrison & Anderson, 1999). This allowed keeping the teaching/learning experience closer to the “normal” experience in schools or universities, since the notion of class and the teacher were re-introduced into DE (Bernath & Rubin, 1998; Hiltz, 1995). The changes were also welcomed by many distance educators, since the previous lack of responsive interaction was perceived as a major educational handicap. However, it soon became apparent that the advantage came with eroding DE’s potential for scale economics (SE), which is a major selling point for DE. One can see this by adapting the cost equation to the situation of online learning (Fig. 2):

Fig. 2
figure 2

Total costs in online education. (Note: Introducing the classroom changes TC into a step function. The larger the class size, the lower the rate of increase in total costs)

Full size image

TC(N) = F + [N/G]*SV + V*N where G indicates group or class size and SV the semivariable costs, incurred when an additional class has to be opened.

After a regrouping of terms:

$$ \textrm{TC}\left(\textrm{N}\right)=\textrm{F}+\left[\textrm{SV}/\textrm{G}+\textrm{V}\right]\ast \textrm{N} $$

Since SV/G > 0, it is obvious that online DE raises unit costs:

$$ \textrm{AC}\left(\textrm{N}\right)=\textrm{F}/\textrm{N}+{\textrm{V}}^0 $$

where V0 = SV/G + V.

Since V0 > V, the online form of DE shifts the break-even point further to the right (i.e., you need, other things being equal, more students to break even) (Thomas Hülsmann, 2016, p. 48).

The lower the cost of the teacher (SV) and the larger the acceptable class size (G), the higher the potential for scale economies. The in-built contradiction of pursuing, on the one hand, the aim of bringing down costs through economies of scale, while on the other hand trying to achieve effectiveness (quality) through improved responsive interaction, was captured in the incompatibility theorem: More scale economies (SE) require restraint in student-teacher interaction (STI) and emphasizing STI erodes SE (Hülsmann, 2014, June, 24).

The realization that online learning with higher levels of STI erodes SE cooled down the initial enthusiasm of the early promoters of a more responsive DE. Anderson had welcomed the new affordances of interaction in the paper on “big and little distance education” (Garrison & Anderson, 1999), but soon adjusted his position in “Getting the mix right” (Anderson, 2003) and, especially, in “Disruptive pedagogies” (Anderson & McGreal, 2012).

Two other findings have further influenced the reassessment of the importance of interaction (and especially STI). The first was the research by Bernard et al., who investigated the relative effectiveness of different forms of interaction (Bernard et al., 2009). Surprisingly, it turned out that among the three interaction formats, SCI was rated first and STI last. This finding was a boon for the proponents of big distance education. Daniel and his team hence were quick to draw the obvious conclusion:

Some important research by Robert Bernard … , explodes the myth about the importance of face-to-face support. They carried out a meta-analysis of hundreds of studies in which distance-education students were treated in different ways. They distinguished three types of interaction: student with content; student with student; and student with teacher. They then analyzed all the studies to find which type of interaction made the greatest difference to student performance when it was increased. The results showed clearly that increasing student–content interaction had much the greatest effect, with student–student interaction coming next and student–teacher interaction last. (Daniel et al., 2009, p. 34)

The second major finding was the Interaction Equivalency Theorem (IET), proposed by Anderson and later by Miyazoe and Anderson (Terry Anderson, 2003; Miyazoe & Anderson, 2010, 2012). The IET states:

Deep and meaningful formal learning is supported as long as one of the three forms of interaction (student–teacher; student–student; student–content) is at a high level. The other two may be offered at minimal levels, or even eliminated, without degrading the educational experience.

High levels of more than one of these three modes will likely provide a more satisfying educational experience, though these experiences may not be as cost or time effective as less interactive learning sequences. (Anderson, 2003, p. 4)

The gist of this theorem resonates with the study mentioned earlier (Bernard et al., 2009). It allows the interpretation that all three forms of interaction can be considered equivalent, and that, to achieve deep learning, only one of the interaction formats needs to be developed at a high level. If all the interaction formats are equivalent, but have very different cost implications, the obvious conclusion is to go for the least costly option. (For the spectrum of teaching options and their respective cost structures, see Fig. 3.) It is quite convenient to find what is educationally the best and turns out to be the cheapest option. Honi soit qui mal y pense.

Fig. 3
figure 3

Spectrum of options

Full size image

Within DE, discussions such as developing the concept of a community of inquiry (CoI) (Arbaugh et al., 2008)) dominated the field and usually proved more innovative than most studies on costs.

The new affordances of ICT created a sort of identity crisis in DE (Guri-Rosenblit, 2008) as indicated by the plethora of new terms and acronyms, such as online learning, e-learning, virtual classrooms or virtual seminars, technology-enhanced learning (TEL), and, more recently, open, distance, and digital education (ODDE).

Open, Distance, and Digital Education (ODDE)

The term ODDE is used here, not to indicate that a new form of DE has taken over from online DE. It is meant to highlight the implications of the digital aspect beyond what was emphasized for online DE. Online DE had inaugurated a new DE teaching option, which reintroduced the use of the class due to the type-c affordances of ICT to sustain responsive interaction at a distance, especially between student and teacher (STI) and among students (SSI). This led to raising the V component in the total cost equation, since time spent in personalized teaching and supporting students is possibly the major factor contributing to V. Increasing V, it was argued, decreases the potential for scale economies (Fig. 3).

STI is, however, not the only factor contributing to V. The other factor is the cost of replication and distribution (RD) of course material. In traditional DE institutions, which often still operate using print material and still distribute it via the postal services, RD remains a significant cost factor. Once the teaching material is digitally captured and the institution is linked to the Internet, these costs fall to close to zero.

The realization that digitally captured knowledge can be made available to all with access to the Internet at minimal costs has huge consequences. It challenges business models in publishing (Open publishing) (Anderson, 2013; Willinsky, 2006) and inaugurated the open educational resources (OER) movement, which dates back to the 2001 MIT publicity coup (MIT OpenCourseWare) to open its archives to the world (at least as far as the world has access to the Internet) (Butcher, 2011).

The step from OERs to massive open online classes (MOOCs) is small. Once ivy-league institutions realized the wealth of the already digitally captured educational resources in their archives, the question arose, why not make them available as courses? The fixed costs of development are zero and, as long as no personalized learner support is promised, V also is zero (V is composed of an STI and an RD factor; the first was zero by design, the second zero due to the fact that RD of digitally captured knowledge over the Internet is zero). Hence:

$$ {\textrm{TC}}_{\textrm{MOOC}}\left(\textrm{N}\right)={\textrm{F}}_{\textrm{MOOC}}+{{\textrm{V}}_{\textrm{MOOC}}}^{\ast }\ \textrm{N}=0 $$

Obviously, this is a somewhat stylized version, since moving a course into the international limelight was, at least in the early days, a publicity coup. Pulling a course out of the archive and turning it into an MOOC required more than “a bit scrubbing” (Weller, 2014, p. 79). Hollands and Tirthali collected data about the costs of developing MOOCs varying between $ 30,000 and $ 300,000 (Hollands & Tirthali, 2014). For instance, to compensate for the lack of personalized support, the inserted quizzes (a more responsive version of the old in-text questions) are instantly evaluated, the data being fed back to the design team to ensure improvements in the next course presentation (Koller, 2012).

If MOOCs can be regarded as “traditional DE reloaded for the digital age” (radicalized in increasing N and reducing V), it comes as no surprise to find it being beset with the same problems (also in a more extreme manner): Completion rates in MOOCs are low, according to some authors below 13% (Ahrache et al., 2013). It turns out that MOOCs require an already seasoned, autonomous, and resilient learner, perhaps even more so than does traditional DE.

The concept that MOOCs were meant to be open (i.e., free to the learner) attracted considerable public attention. In the wake of the GFC, when Higher Education (HE) funding had been cut back at a time when HE was becoming increasingly necessary for a better job, tuition (in the USA and the UK) soared way beyond home prices and the consumer price index (Thomas Hülsmann, 2016, p. 22). Education appeared to be “broken,” unable to serve as a vehicle for upward mobility. In this situation, MOOCs emerged with the promise to rebuild HE in a very American way: through the combined forces of technology and venture capital (Bates, 2015, p. 176). However, the very openness presented a problem for this combination. Being open, MOOCs did not, at least in the early days, produce any substantial streams of revenue.

From HCT1 to HCT2

This gives the opportunity to comment on some changes in the “macroeconomic weather conditions.” Early human capital theory (here labeled HCT1, to distinguish it from a later mutation) was based on the perception that education would raise general productivity and should be seen as an investment. The profitability crisis of the 1970s, however, inaugurated the neoliberal project in major capitalist countries (Todaro & Smith, 2003). Margaret Thatcher famously declared that “there is no society but only individuals” and that “all the money the government can dispose of is taxpayers’ money,” which, everybody knew, was scarce. Consequently, selling off public monopolies such as roads and other infrastructure was turned into opportunities for rent extraction. “This turns the economy into a set of tollbooths as user-fees raised on labor, industry and other non-financial ‘real’ activity” (Hudson, 2012). The user of the highway pays a fee for using it, the ill have to pay for their treatment, and the students for their education. Since social surveys indicated that the educated enjoy higher lifetime earnings than the less educated, students can be expected to take out loans for their education (Hülsmann, 2011; Spraul, 2006), a development here referred to as HCT2. The perception of HCT in education had changed: While earlier it had been seen mainly about enhancing societal productivity (by reducing, for instance, overhead costs coming with training), the changed perception now regarded education as a center of profit itself.

In education, this shift led to various attempts to participate in the education market. An early attempt was to open for profit universities. This proved not to be too successful (Hanna, 2003, p. 69). More promising were the inroads in the HE market due to the proliferation of ICT in education. Universities were among the first to be connected to the Internet. They found the emerging web-based learning management systems (LMS) useful, also on-campus.

When, in the wake of the GFC, much of the HE funding was slashed, universities looked for diversifying their income streams. Educational technology (EdTech) providers already played a prominent role in various services. It seemed a win-win situation for HE and EdTech providers to join forces. In some cases, the entire online management programs (OMP) were outsourced to EdTech companies, which allowed increasing student intake. The added revenue (fees) were split between the university and the OMP provider. At times, the bigger share (60%) goes to the EdTech provider, which still helps the HE institution since it gets 40% instead of, as before, nothing (Carey, 2019, February 6).

After this short tour d’horizon of DE costs structures, the potential of DE in terms of cost-efficiency should be plausible. However, the rising tuition fees in many countries suggest that, in spite of bringing the unit costs down, the cost-efficiency gains were not handed down to the students. Instead, devolving costs to the students compensating for reduced state funding, often increased tuition, and in the wake, student debts.

Modern Money Theory (MMT)

Earlier in this chapter, it was argued that the present three crises require reconsidering the understanding of money and state. In the next section, the history of DE with a focus on its economic aspects was revisited. It was shown that the costs and economics of DE had been framed by the neoclassical economic paradigm predicated on an implicit understanding of money and state. According to this paradigm, the only money the state disposes of is the scarce taxpayers’ money. In this frame, it seemed plausible to devolve educational cost to the students who, after all, are supposed to benefit from it in terms of higher lifetime earnings.

MMT has a different understanding of money and state. For MMT, any analysis that confines the costs and economics of DE to the perspective – anticipating MMT terminology – of the users of the currency is insufficient. Education, like health and infrastructure, has traditionally been under the purview of government, which, in MMT terms, is the issuer of the currency. Since money in the chartalist MMT tradition is a creature of law and, by extension, the state (Knapp, 1905), money is neither scarce, nor is it necessarily taxpayers’ money. Therefore, MMT would arrive at different consequences for the costs and economics of DE by (i) exposing its traditional microeconomic focus as too narrow and (ii) devolving educational costs to the learner as misguided.

MMT Basics

Inspecting the understanding of MMT, it shows three important features:

  1. 1.

    Money is a creation of law (and by extension of the state).

  2. 2.

    The acceptance of money is driven by taxes.

  3. 3.

    Money can be understood as tax receipt (tax credit).

The first point situates MMT in the chartalist tradition. Chartalism sees money as a creature of law (Knapp, 1905). MMT starts with a distinction between the issuer and the users of the national currency (Mitchell et al., 2019). All citizens, all private sector institutions, and even local governments are users of the currency. The sole legal issuer of the currency is the government. MMT usually refers to the government (or “consolidated government”) as the ministry of finance and the central bank together (in the USA, the Treasury and the FED).

In the chartalist tradition, money has no intrinsic value and is often referred to as fiat money. That distinguished it from an earlier metallist tradition, which saw money as being tied to gold or silver. Since the end of the gold standard (1971), the links to the metallist tradition were severed inaugurating the modern fiat money economy. The question, how governments were able to get an intrinsically valueless currency accepted, leads to point two of the list above: The acceptance is driven by taxation. The government holds two monopolies: the monopoly of power and the monopoly to issue the national currency. By imposing taxes payable in its own currency, the government creates a demand for its (intrinsically valueless) currency. For example, between 1755 and 1774 the state of Virginia issued its own currency. Having to pay taxes in Virginia, Pounds (₤VA) colonists were forced to offer part of their resources (most importantly labor) to earn the state’s currency. Grubb reported that the tax money “redeemed” was literally burned (Grubb, 2015; Wray, 2019). Since all colonists had to get hold of the currency to pay their taxes, the intrinsically valueless ₤VA notes were accepted as general means of payment.

The (general) stylized story that MMT tells is that first the government agents (soldiers, police officers, judges, administrators, etc.) were paid in the government’s currency. They would accept the government’s currency, since they could buy anything with the government’s money such as shoes and shirts and bread, because the shoemaker, the tailor, and the baker also urgently needed to get hold of the government’s currency for paying their taxes. In this way, “taxes drive money” (Mitchell et al., 2019 p. 137) or, more precisely: Taxes drive the acceptance of the currency. Note that, by setting the salaries of its agents, the government defines the reference frame of labor costs. The salaries have to be sufficiently high for the government’s agents to reproduce their labor power or, better, high enough to ensure their loyalty. Other salaries will eventually judder in place accordingly (Höfgen, 2020, p. 105). Beyond getting the government’s currency accepted, there are further functions taxes can support: They can be a tool in demand management, they can be used to battle inequality, or to influence expenditure or behavioral decisions; financing the government’s budget is not among them (Bell, 2000; Höfgen, 2020, p. 121; Ruml, 1946).

This brings us to point three. Since all citizens need to pay taxes, and one can pay taxes only in the government’s currency, money can be seen as tax receipt or tax credit. Note the double inversion in which MMT differs from the mainstream view: First, in the mainstream account, the government needs the taxpayers’ money to “finance” its budget, while for MMT, the citizens need the government’s currency to pay their taxes. Since the government is the currency issuer (creator of fiat money), it makes no sense to assume that it is after the tax payers’ money. There is obviously something else it wants: It wants some of the people’s resources, most importantly their labor! To put it simply: The government creates fiat money (by issuing money icons, e.g., notes, or simply keying numbers in authorized spreadsheets); imposing taxes in that currency incentivizes a maze of economic activities (e.g., producing commodities, providing services), which in turn underpin the fiat money with value. Second, while in the mainstream, account taxing precedes (government) spending and even sets the limit for its budget, in MMT, spending has to precede taxing (Kelton, 2020). Moreover, since the government as currency issuer faces no inherent limit in creating fiat money, it follows what one may call the “main theorem of MMT”: “A sovereign government does not face any financial constraints. It always can buy anything on offer in its own currency” (Mitchell et al., 2019 p. 13). Obviously, this does only apply for governments which did not tie their currency to another country’s currency or a commodity (e.g., gold) or which are indebted in a foreign currency. Hence, there is a spectrum of sovereignty. The USA enjoys high monetary sovereignty, the UK, Australia also, while Zimbabwe has little monetary sovereignty. Countries of the European Monetary Union (EMU) are sitting somewhere in the middle since they use a currency not controlled by a national government.

The MMT main theorem is accompanied by a corollary: that even sovereign governments face real resource constraints, i.e., constraints in labor, natural resources, or knowledge resources. But: “Anything what we actually can do we can afford” (Tooze, 2021, p. 22, quoting Keynes’ 1942 BBC Address). The limit is not: “Can we afford it?” The limit lies in the actually available real resources. Spending into an economy, where all resources are already activated, leads to inflation. Involuntary unemployment, however, is a clear indicator of unused capacity.

Two dangers are usually invoked: the mounting public debt and the (alleged) likelihood of inflation. Inflation describes the process when average prices rise beyond average wages. MMT rejects the monetarist claims that inflation is driven by the expansion of money supply (cf. “quantity theory of money”) both on theoretical and empirical grounds. Instead, it sees a strong link between inflation and unit labor costs (Flassbeck et al., 2020, pp. 72, 94). Only increasing wages beyond the level of productivity increases under conditions of full employment leads to inflation. Central banks observing such developments usually can reduce demand from labor by increasing interest rates to manage inflation. The mounting of public debt leads us to the next section where sectoral analysis will throw a new light on what public debt means.

Sector Analysis and the Mechanics of Balance Sheets

The government issues fiat money and spends it into the economy. Hence, the money in the nongovernment sector increases while the money in the government sector decreases. However, “money flow” is a metaphor too close to the traditional understanding of money as a thing, which “changes hands” or flows from one bucket to another. It is more appropriate to conceive the monetary system as a set of related balance sheets, in which claims accumulated in one balance are exactly mirrored by liabilities in someone else’s balance. In a closed economic system (and the world as a whole is a closed economic system), the net financial balance is at each point in time zero (so are the net financial liabilities) (Flassbeck et al., 2017, p. 4). It also shows that the world as a whole cannot “save” (acquire net financial assets) for the future: Wealth consists of net financial assets and tangible assets. Since the world’s net financial assets are always zero, the wealth of the world consists of its tangible assets (its capital stock). This shows how misguided austerity policies are from a macroeconomic vantage point when it tries building financial wealth at the expense of the maintenance of the capital stock.

Godley’s “one equation model of the world” (Kelton, 2019, p. 105) states:

$$ \textrm{Government}\ \textrm{financial}\ \textrm{balance}+\textrm{Nongovernment}\ \textrm{financial}\ \textrm{balance}=\textrm{Zero} $$

This means that a deficit in the government’s financial balance implies a surplus in the nongovernment sector’s financial balance.

Usually economists partition an economy in three sectors, the public sector, the private sector, and the foreign sector. If one ignores for a moment the foreign sector (it may be marginal or balanced), Godley’s equation would read:

$$ \textrm{Government}\ \textrm{financial}\ \textrm{balance}+\textrm{Private}\ \textrm{sector}\ \textrm{financial}\ \textrm{balance}=\textrm{Zero} $$

This is equivalent to saying that the government’s deficit is equal to a surplus, the private sector surplus. This is a major MMT punchline: “Their red ink is our black ink” (Kelton, 2020, p. 101 ff) (Fig. 4).

Fig. 4
figure 4

Sectoral financial balances in US economy 1990–2017. (Note: The diagram illustrates how the private sector financial balance and the government sector financial balance mirror each other. By definition, the three balances must net to zero

Full size image

During the Clinton administration (1993–2001), the government reduced the deficit (and even achieved a surplus). However, at the expense of a massive private sector deficit.)

Misled by the (wrong) analogy of the government’s balance and a household budget, it is widely preferred to have a surplus not only in the private sector but also in the public sector. This is, in principle, possible (for some countries) if the foreign sector balance (import minus export) is negative, i.e., exports exceed imports. Germany is a point in case. However, sector analysis shows that this is not a generalizable option: A surplus country requires a debtor country since the accounting system matches each surplus with a deficit, each credit with a debt. It is not possible to wipe out all debts (liabilities) without wiping out all credits (claims) at their same time. (Note that, in a “normal” market economy, businesses borrow to invest and pay back the credits out of the profits.) Since meanwhile the business sector balance sheets (in the USA or Germany) also show surpluses, the governments have to pick up the debtor role.

Fiat Money

In a modern economy, we distinguish between two monetary circuits. The circuit of reserves connects the central bank with the commercial banks, while the circuit of money of account connects the commercial banks with their customers. Each bank has an account at the central bank, and each household or business has at least one account at a commercial bank.

The creation of fiat money is done on two levels: on the level of commercial banking through lending and on the government level through spending into the economy (deficit spending).

Until recently, commercial banks were seen as mere intermediaries, whose lending depended on what savers have deposited. Meanwhile, it is widely accepted that “loans create deposits” (McLeay et al., 2014; Werner, 2014). In the words of an economist of the European Central Bank (ECB): Money is created since the commercial bank pays with a claim against itself (liability). Such liabilities of commercial banks are counted as part of the money supply (Biswanger, 2012, p. 29). (If an apple farmer wants a credit, the bank simply keys a number into his account. The farmer, on the other hand, has to produce real apples and sell them to service the loan.) In reality, money creation is driven more by commercial lending than by central bank activities (Douglas & Raudla, 2020, p. 9). In principle, this creation of fiat money is a flexible way of providing the economy with the needed liquidity. If all works well, the money is used productively and incentivizes activities in the real economy rather than being used for speculation.

Money creation on the government level involves the central bank and the ministry of finance. MMT considers the central bank as being “owned” by the government and refers to the central bank and the ministry of finance as “consolidated government.” If the government needs money to spend into the economy, it advises the central bank to credit the account the reserves needed. The central bank does so while entering the same amount as liability in its own account. Say, the government wants to construct a harbor. Then the ministry of finance convenes a consortium of companies to do so. It advises the central bank to create the reserves necessary to equip the banks representing the consortium with the reserves necessary to enable them to credit the companies with the necessary money of account. In the end, the government gets the harbor, the companies get their profits, and the workers get their salaries. There is nothing here which has to be paid back … unless (for questionable reasons) the government is not allowed to run a deficit. MMT considers this as self-imposed constraint.

Generally, this forces the governments to sell bonds to the banking sector. Due to the deficit (!), the banking sector can do that and, since it involves simply swapping the reserves the banks keep at their central bank accounts with bonds (i.e., interest-bearing reserves), is very interested to do that.

As long as money creation (by commercial banks or the central bank) incentivizes productive activities in the real economy to realize the government’s democratic mandate, there is little danger of inflation. However, while governments can create money, they cannot control its use. Often it is not invested in the real economy but for the buyback of shares or in the FIRE sector (FIRE = Finance Insurance Real Estates) leading to a form of “nightmare MMT” where the expansion of the money supply is, via internal mechanisms of our economic arrangements, handed through to the billionaire class (Hudson et al., 2020b, April 10). To prevent that, monetary and fiscal policies have to be coordinated.

Normative Extensions

MMT refers to itself as a descriptive theory. It claims to provide a superior description of the mechanisms of the monetary system in the USA (and nations of similar monetary sovereignty) (S. T. Fullwiler, September 2010 (edited April 2011); E. Tymoigne, 2016; É. Tymoigne & Wray, 2013). This understanding provides MMT with a lens, which allows identifying additional policy space. Government spending is not constrained by tax revenues (or the selling of bonds), but by inflation signaling real resource constraints. This creates additional policy space. How this space is used poses normative decisions and cannot be deduced from the descriptive kernel of the theory.

MMT sees economics as “the study of creation and distribution of society’s resources” (Mitchell et al., 2019 p. 7). It is not seen as a natural science or a form of applied mathematics, the logic of which renders the negotiation of societal goals irrelevant. MMT researchers therefore explore the viability of their normative preferences in the light of the MMT framework. The first issue is unemployment, which they proposed to address by a general Job Guarantee (JG). The idea here is simple: While at present the economy operates with a buffer stock of unemployed people, the JG proposes to operate with a buffer stock of employed people. MMT proposes a centrally funded but locally administered “public service employment” (PSE) (Randall Wray & Kelton, 2018; Tcherneva, 2018). Another issue is the Green New Deal (GND) which relates to the climate crisis (Nersisyan & Randall, 2019). In relation to education, the issue of student debt cancellation has been explored (Fullwiler et al., 2018, p. 50). It was found that student debt cancellation would be feasible and, instead of being an unsustainable burden, would act as a positive economic stimulus (since it would free income to spend back into the economy). The impact on inflation was found to be negligible (Fullwiler et al., 2018, p. 50).

Covid, Conclusions, and Caveats

Covid: This chapter is written under the conditions of covid-19. The pandemic led to major government interventions, some pertaining to the economy as a whole, others directly impacting education. Monetary sovereign governments were able to impose lockdowns, intentionally putting the economy in artificial coma, while keeping it on life support by launching massive stabilizing programs (Bibow, 2020).

With respect to education, the pandemic-imposed “social distancing” should have provided a through pass for “distance education.” And in higher education, DE worked reasonably well; since universities were reasonably digitally connected, students usually owned the necessary mobile devices and were competent in operating them (Dolch et al., 2021). In schools, the pandemic-induced precipitated shift to DE (referred to as Emergency Remote Teaching (ERT) (Hodges et al., 2020)) worked less well. Due to the varying quality of the Internet infrastructure, basic preconditions for online teaching were not in place. Pupils did not own the necessary end devices, nor had they easy access to them. What they had varied according to the socioeconomic status of the family. Teachers have to deal with devices of varying quality, often lacking interoperability. Both problems show that ERT was facing problems beyond the reach of individual institutions.

Conclusion: The first part of this chapter sketched the costs and economics of DE. Costing is part of microeconomics. MMT, as a macroeconomic discipline, has little to add here. The insights in the cost-efficiency of DE remain valid. Financing education is part of macroeconomics. It is here that MMT makes a difference. When discussing financing, distance educators find it necessary to call for redistributive taxation (e.g., Rumble, 2007) or want to bring in private capital (e.g., Daniel et al., 2006). MMT tells us that monetary sovereign governments do not have to rely on taxes for financing their goals as long as it has the required resources. Money is debt and is created either as private or public debt. In the first case, it is created through commercial lending, in the second case as government “debt.” The change of macroeconomic weather conditions (cf. HCT 2) gave preference to private rather than public debt. Understanding sectoral analysis changes the perception of public debt (cf. Kelton’s “their red ink is our black ink”). It is less intimidating than private debt, since the (consolidated) government is in a much different debtor position than an individual. Moreover, MMT identifies policy space, beyond tax income and below the inflation threshold, where some policy agenda can be realized: free (or subsidized) education, health, and infrastructure. It can be used for implementing the initial mission of DE: to widen access to education, including HE, beyond its traditional audience. In contrast, under the neoliberal preference for commercial lending (private debt), the real economy of production and consumption is surrounded by a tightening network of tollbooths where surplus is siphoned off in form of interests and rents to the FIRE sector. The dynamics of compound interest threatens “killing the host,” i.e., the real economy (Hudson, 2015, 2017). Student debt is one of such tollbooths.

Hence, MMT shows that it is possible to widen access to education (including HE) without devolving the costs to students, nudging them toward debt peonage. This allows DE to relax its efficiency drive and build in more resilience in the its system, sadly missed in times of emergency.

Caveats. First, MMT is based on the notion of monetary sovereignty. Not only is there a spectrum, but also a hierarchy of monetary sovereignty with the US$ in a position of dominance. This limits what can be transferred. Nevertheless, understanding modern money systems takes some of the TINA sting out of the “there is no money” austerity mantra. Second, the MMT insistence that taxes only serve to ensure the currency acceptance underplays their importance in preventing state capture (e.g., regulatory capture). Quantitative Easing illustrates that spending money into the economy does not prevent it ending in the coffers of the 1% (Hudson et al., 2020a). Third, admittedly this presentation of MMT lives not up to the depth and the scope the discussion of the topic merits. For more, the reader is referred to (“Modern monetary theory and its critics,” 2019, October, 11). The more mathematically inclined may download the (free) Minsky software and view S. Keen’s tutorials on how to use it to explore some MMT aspects (https://www.youtube.com/watch?v=Dt4thL3eToU).

Cross-References