Godley Tables

Minsky ’s double-entry bookkeeping tables are named after Wynne Godley, for three reasons:

• Godley, in collaboration with Francis Cripps, was the originator of the concept of using double-entry bookkeeping tables to ensure stock-flow consistent modelling;
• I spent six very pleasant months at the Levy Institute in 2000, writing the first edition of Debunking Economics while on sabbatical leave, and I learnt a great deal from interacting with Wynne at that time, including the crucial role of double-entry bookkeeping in ensuring stock-flow consistency; and
• Because non-Neoclassical economics needs to preserve the names of its heroes. If we leave history to the victors—which, in the sad case of economics, means leaving it to Neoclassical economists—then the names of our heroes will be forgotten. Hence the name of Minsky itself, Godley Tables for our double-entry bookkeeping tool, and—if I can raise further development funding after our £200,000 grant from Friends Provident Foundation runs out— Moore Tables to show the macroeconomics of inter-sectoral flows, to honour Basil Moore (Moore 1979, 1988b).

Godley Tables in Minsky differ from the flow matrix tables in Godley’s own work. Whereas both the rows and columns in his tables summed to zero “on the principle that every flow comes from somewhere and goes somewhere” (Godley 1999, p. 394), the rows in a Godley Table sum to zero, but the sum of the columns adds up all the flows into and out of a given account, and therefore tells you the rate of change of the account the column represents.

Therefore, when you fill out the rows in a Godley Table, you are actually building a set of differential equations with which to model an integrated financial system. The rule enforced by the Godley Table, that each row must sum to zero, ensure that these differential equations are stock-flow consistent.

7.1 Creating a Godley Table

There are 2 ways to insert a Godley Table onto the canvas: click on the Godley Table icon on the toolbar, and then click on the canvas where you wish to place it; or choose “Godley Table” from the main menu item “Insert”.

When you first create a Godley Table, you get a bank icon on the canvas—see Figure 106.

Figure 109: A blank Godley Table icon

Double-click on the icon, or click the right-mouse button and choose “Open Godley Table”, and Figure 107 will appear, inside a new Window.

Figure 110: A blank Godley Table opened in an editing window

The top row labels each account as either an Asset, a Liability, or Equity—the difference between Assets and Liabilities. The final column, labelled applies the “golden rule of accounting”, that Assets minus Liabilities minus Equity equals zero, to each row in the table. A−L−E,

Immediately below this line has buttons to add or delete columns. There is one set of buttons for each of Asset, Liability and also the Equity columns, if you enable multiple equity columns from the Options menu (if you don’t, there are no buttons for the Equity column). The + key adds a new – column to the right, the key deletes the current column, and the arrow keys move the selected column one position to the left ← or right → .

The third line starts with the top left cell in the table, which notes that the columns are “Stock variables”, while the rows are flows between these stock variables. The columns to the right are where you type the names of the accounts (the down-triangle icon is discussed later). The row below this shows the initial conditions for the accounts—the amount of money in each account when a simulation commences—which must also follow the rule that the numerical sum of these conditions must be zero. At the left-hand end of this line is a plus key, which A−L−E creates the first row. Once you have done this, plus, minus, up and down symbols appear to allow you to add and delete rows, and move them up and down.

While there can be numerous entries in a row, the norm is two, which must sum to zero according to the rule A A A −L L −E , which is checked by the column. The entries are symbolic: words, not numbers. These words can include the formatting tricks discussed in the first chapter—subscripts, superscripts, grouped text and Greek letters—and if TL ALA ELAE= 0 A−L−ES you can have 0.7 × S in one column and 0.3 × S in another. AS Now let’s use AS Minsky to build the simplest possible models of a monetary economy, starting with a AS

Now let’s use Minsky to build the simplest possible models of a monetary economy, starting with a model of a pure credit economy in which all money is created by bank loans.

7.2 The simplest possible monetary model of a pure credit economy

Figure 108 shows a simple model with credit (bank-created) money only, with six flows: lending to firms, interest payments, debt repayment, wages, workers’ consumption, and bank purchases from firms.

Figure 111: A simple Godley Table

Minsky takes these entries and creates a set of ordinary differential equations, which you can see either by clicking on the Equations Tab, or by choosing “Export Canvas” from the File menu, and then choosing the file type to be LaTeX (*.tex). Equation (43) shows the differential equations for Figure 108.

Though this is quite a simple “toy” model, the same process enables huge, detailed models of the financial system to be built, with complete confidence that these equations are stock-flow consistent.

Once you have made entries in a Godley Table, the Godley icon on the canvas changes to show the flows as inputs on the left-hand side, and the stocks as outputs on the bottom:

Figure 112: A Godley Table after stocks and flows have been defined

You can also alter the view so that you see the actual Godley Table on the canvas. Choose “Editor Mode” from the right-click menu, and the table will display as shown in Figure 110.

Figure 113: The Godley Table in Editor Mode

As the name of this display mode implies, you can edit the table here rather than in a separate window, but you have to activate the row and column buttons that are shown automatically in the separate window. You can also turn on showing the stocks and flows attached to the table via the

“Display Variables” option on the right-click menu—see Figure 111 (I have also added a title to the Table, using the right-click menu option “Title”).

Figure 114: The Table showing editing buttons and the Stock and Flow widgets

One thing Russell Standish and I have focused upon in designing Minsky is enabling quality documentation of a model by Minsky itself. This includes the capacity to export a Table in either CSV or LaTeX format, via the “Export to File” option on the right-click menu (and also on the file menu from within a Godley Table window). Figure 112 shows the LaTeX output for the Table in Figure 111.

Figure 115: A screenshot of the LaTeX rendition of a Godley Table

7.3 Defining the flow elements of a Godley Table

The Godley Table itself clarifies issues in monetary economics, even without a simulation—especially when it is linked to other Godley Tables in an integrated model. But it also makes the modelling of monetary dynamics so much easier than with flowchart programs, as noted above.

To turn a Godley Table into a model, the flows in it have to be defined on the canvas itself, using the same flowchart logic as in the previous section. There are two ways to get the stock and flow variables in a Godley Table onto the canvas: individually, by right-clicking on the flow or stock variables attached to either the icon (Figure 109) or the table (Figure 111); or by right-clicking on the

table and choosing “Copy Flow Variables” and “Copy Stock Variables” commands which copies all the relevant variables at once. Figure 113 shows the canvas after they have all been copied.[34]

Figure 116: All the stock and flow variables from a Godley Table copied to the canvas

How you define a model is up to you, but you can only define it using the variables and stocks you currently have, or transformations of them—so if you want to define investment flows in a model as a function of capacity utilization, for example, or the wage level as a function of the level of employment, then you need to add those variables to your system. Here I’ll just demonstrate defining a model from the elements of the Godley Table itself.

The simplest flow to define is InterestF , which is the rate of interest on loans multiplied by the current level of Loans. In Figure 114 I add a new parameter, rLoans__F , for the rate of interest on loans to firms, multiply that by the stock variable Loans, and this defines InterestF (I have also shrunk the Godley Table using its bounding box arrows).

Figure 117: Defining interest payments (without hitting the “Recalc” button before exporting the figure)

To define the other flows, I use the mechanism I explained earlier in these models—the first-order time lag. As well as being useful to define a lagged variable, such as lagged inflation as a function of the actual rate of inflation, it can be used to explain a flow as a time-based response to a stock. For example, the level of repayment of existing debt will be roughly proportional to the current level of loans—it can be a large proportion or a small one, but there will be some proportionality. This could be done using a simple scalar—say, for example 10%, so that 10% of loans are repaid every year, as in Figure 115.

Figure 118: Repayment modelled using a repayment rate (after hitting the “Recalc” button)

I prefer to use a time constant instead, because then the value of the time constant is easily understood in terms of the time dimension of the model. If I give that time constant a value of 10, as in Figure 116, then I get the same numerical result as in Figure 115, but the number 10 stands for the number of years it would take to reduce the debt to zero if this rate of repayment were sustained.

Figure 119: Using a time constant instead for the rate of debt repayment

A similar definition for the rate of new lending tells you how long this rate of lending would take to double the debt—see Figure 117, where I’ve also copied the model’s parameters to the top of the canvas, where they can be varied easily during a simulation.

Figure 120: Lending also with a time constant, plus copying parameters into a "control panel"

This simple model, without any physical output or price component, needs a definition of GDP as well in terms of financial flows only. We can’t just add up consumption and investment using this model, because the only non-financial flows in it are Wages as the income of workers, and consumption by workers ( ConsumeW ) and bankers ( BuyB ): there’s no definition of profit as the income of capitalists, nor is there any definition of investment—which involves capitalists buying from other capitalists, and is therefore subsumed within the single column for the Firm sector. So,

given how this simple model is constructed, investment—as well as profit, and consumption by capitalists—doesn’t appear at all, and it therefore has to be inferred as a residual.

My approach in these simple models (without an integrated model of the physical economy as well) has been to take a leaf out of Das Kapital —specifically, Volume II, Chapter 7, “The Turnover Time and the Number of Turnovers”:[35]

We have seen that the entire time of turnover of a given capital is equal to the sum of its time of circulation and its time of production. It is the period of time from the moment of the advance of capital-value in a definite form to the return of the functioning capital-value in the same form. (Marx and Engels 1885)

In this model, GDP is derived from the amount of money in the Firm sector, and its turnover rate:

This equation, in flowchart form, is highlighted in grey in Figure 118.

Figure 121: GDP as the turnover of money per year in the firm sector

With GDP defined this way, inter-firm spending (which in this simple model, includes investment and consumption by capitalists, since I haven’t separated out capitalists as a different financial entity to Firms in this model) is the residual between GDP and wages. This residual includes profits, dividends, etc.—again, aspects of a capitalist system that you could explicitly model in a more elaborate model. I also use a very simple assumption to determine wages: I make the distribution of income a parameter, so that of GDP goes to workers and goes to capitalists as gross profit, ω% (1 −ω)%

>

with this minus interest payments being net profit (in an integrated physical-monetary economy model, wage determination would be driven by bargaining power, as in the Goodwin model).

This leaves consumption by workers and bankers to be defined. You could, of course, make a “Kaleckian” assumption that workers simply consume their wages, and equivalently, that bankers spend their interest income. For the sake of illustration, I’ll do that first (in Figure 119), and compare the results to a model with time constants, based on the amount of money in the workers’ and bankers’ accounts.

Figure 122: "Kaleckian" assumptions on consumption by workers and bankers

This Kaleckian assumption effectively making workers and bankers passive parts of the system, rather than active parts: whatever they receive as Wages ($216/Year) or Interest ($5.50/Year) goes out as consumption. On the other hand, if you base spending upon the amounts in their bank accounts, divided by time constants that reflect the fact that workers are living close to “hand to mouth” whereas bankers have large buffers compared to their spending, then you have a small time constant for Workers and a large time constant for Bankers. Changing the distribution of income between workers and bankers will therefore change the amount turning up in the Firms account, thus changing GDP.

Figure 123: Consumption based on time constants

Figure 120 also shows the advantages of dividing by a time constant to define a flow, rather than multiplying by an equivalent constant. The value for the time constant tells you how long, in fractions of a year, that the social class could consume before running out of money: 1/25th of a year for Workers (otherwise known as a fortnight), 2.5 years for Bankers. The size of the time constant is readily interpreted as an indicator of the relative income and wealth of the two social classes.

Figure 121 illustrates the impact of varying the time constants in the model. If the time constant for lending is smaller than that for repayment, then there is net debt and money creation by the banking sector, and GDP rises. If repayment is faster than new lending, then there is net money destruction and GDP falls. Changes in the workers’ time constant have more impact than changing that for bankers—workers spend their accounts much more quickly than bankers, because they have to. So though their bank accounts are the same size in Figure 121, workers generate far more spending than do bankers ($250/Year versus $4/Year).

Figure 124: Varying time constants for lending, repayment, worker & banker consumption

By using the “Editor Mode” display of the Godley Table, and choosing “Godley Table Show Values” from the Preferences form of the Options main menu (see Figure 122), you can see the amounts passing between the accounts in this model in the Godley Table itself—see Figure 123.

Figure 125: The Preferences form from the Options menu

Figure 126: Editor Mode display with numerical values shown in the Godley Table

The last step in putting together a comprehensive model is to show the financial system from the point of view of Firms and Workers, as well as from the Banking sector. To do this, insert two more Godley Tables on the canvas, label one Firms and the other Workers , and then use the down-arrow on the columns ( ) to search for Liabilities that haven’t yet been defined as Assets, and vice versa. The Firm sector has one Asset—its deposit account Firms —and one Liability— Loans . When these are added to its Godley Table, Minsky automatically fills in the rows where there are already operations on both accounts ( LendF and RepayF ), while leaving those where there is an operation on one but not the other unbalanced: the sum of shows the flow that hasn’t yet been allocated to an account—see Figure 124. A−L−E

Figure 127: The Firm sector's Godley Table with Assets & Liabilities added

To fully specify the model, you need to define an Equity column for the Table. I used FirmE as the name for “Firm Equity” and made the matching entry needed so that on every row— see Figure 125. In this simple model, all those entries go in the Equity column, but that isn’t necessarily the case in a more complex model. You might, for example, have A−L−E= 0 CashW as an asset of the Workers, so that withdrawing money from the Banking sector reduced the Workers’ deposit account ( Workers ) and increased their cash account CashW , without altering their Equity.

Figure 128: The model from the Firm sector's point of view

Figure 126 shows the complete model, with all accounts recorded in the respective Godley Tables.

Figure 129: The complete model, which is still very simple, with financial dynamics only

7.4 Getting creative with Godley: Modelling the pandemic

Godley Tables were developed to enable monetary modelling, but that’s not all they can do. The basic idea that Wynne Godley himself based his modelling on—that “all financial flows start somewhere and end somewhere”—makes it feasible to use Godley Tables to model processes where there is an exclusive transfer from one location to another.

One highly relevant instance of this is the Covid-19 pandemic. You either have Covid or you don’t; you’re either in hospital, or you aren’t: the various positions someone can be in with respect to the disease are mutually exclusive categories, and if you enter one category, you leave another.

I used this to build a simple model of a pandemic—known as a SIR model, where the acronym stands for SusceptibleInfectedRecovered—which I’ll explain in a moment. Subsequently, a supporter of

mine on Patreon, Tyrone Keynes, developed an incredibly sophisticated model of the pandemic, which really shows the power of Minsky and its Godley Tables in system dynamics modelling. I’ll showcase Tyrone’s model after I’ve explained the simplest possible model.

In the simplest possible model, you’re either Susceptible but not Infected, Infected, or Recovered; in a slightly more complicated but highly relevant model, there’s a 4th option—Dead.

This basic pattern is easily implemented in a Godley Table. Create a Table with 5 columns: N for the population—which is assumed to be constant, since the human population grows much more slowly than the disease spreads, and make this the Table’s sole Asset. Record as Liabilities, S for susceptible—which initially is almost everyone in the population; I for Infected—which is equal to the population minus those susceptible; and D for Dead—which starts at zero. Show R for Recovered as the model’s Equity—which also starts at zero.

Then define flows between each of the possibilities: Catch for getting the disease; Recover for Recovering; and Die for dying. Put in initial conditions—which, as in any monetary model, must sum to zero using the A-L-E rule—and you have the structure for a basic model.

Figure 130: A basic SIRD model in Minsky

Minsky automatically generates the differential equations for this model:

Because the pandemic grows so much faster than the population, can be treated as a constant.[36] That enables the N on the left hand side of the equation to be cancelled: N

The equation for the percentage rate of growth of infections is then:

The equations for recovery and death are defined using simple constants, but to take advantage of the system dynamics framework, I’m using a “time constant” for the duration of the disease. In Figure 128 I have set the infectious parameter to 0.5, D to 14 days, and the fatality rate to 5%. The characteristic “pandemic wave” with which we have become all too familiar since 2020 DatD l β τ shows up in the plots.

Figure 131: A simple SIRD model of a pandemic in Minsky

This simple model is more representative of a disease in an animal population where there is no coordinated response to the pandemic itself. Of course, in our human civilization, a huge diversity of responses were tried, to varying degrees of both success and compliance: “social distancing”, mask mandates, lockdowns, travel restrictions, etc. This is where Tyrone Keynes’s[37] model comes in.

It has not a mere 5 system states, as in the model in Figure 128, but 25 states, and numerous policy settings that can be varied dynamically as the model runs. The imposition and premature release of these controls generates the characteristic multiple waves that we have seen in our real world Covid experience, strongly suggesting that it was this very on-again, off-again public policy process that generated the waves. As the pandemic and complex systems control expert Yaneer Bar-Yam argues on https://www.endcoronavirus.org/, the only way to successfully fight a pandemic is to react too soon, to over-react, and to not remove controls until cases have fallen to zero.

A major motivation in designing Minsky was to produce a tool by which policymakers could learn lessons like these “in silicon” before confronting them in the real world. Tyrone’s model is a superb example of how models like this could be constructed in Minsky , so that the policymakers could learn how to manage real world crises before they occur in the real world.

Figure 132: Tyrone Keynes's pandemic model with varying policy controls over time

I hope that’s enough background to enable you to use Godley Tables in your own modelling. Now let’s use Minsky to show why, when it comes to money, Paul Krugman doesn’t know what he’s talking about.

7.5 A Keen Rant : Using Minsky to Revisit the Keen-Krugman Debate

Almost a decade ago now, Paul Krugman gave me a birthday present, by citing me in his New York Times column on March 27th , 2012.[38]

Entitled “Minsky and Methodology (Wonkish)”, the post began as any birthday present should—it was very nicely wrapped:

Figure 133: The opening to Krugman's first post

Unfortunately, once I opened the present, it was all downhill. He wrote a series of seven posts,[39] ending with “Oh My, Steve Keen Edition”, whose final line was “Nick [Rowe] uses a four-letter word to describe this; I can’t, because this is the Times.”

In between the nice introduction and the derogatory denouement, there was something that is far too rare in economics today, a “debate” between opposing schools of thought in economics over a fundamental issue. I put “debate” in inverted commas because we never spoke, and while I read his posts, he didn’t read mine.[40] But the juxtaposition of opposing views was something that rarely happens in economics, so in that sense, it qualifies as a debate.

The topic of the debate was “Do banks, debt and money matter in macroeconomics?” Krugman’s position was “No” back then, and it’s still “No” today: in the 2021 promotional video for his Masterclass on economics,[41] he says “It’s about people. It’s not about money”.

39Krugman’s posts on me https://krugman.blogs.nytimes.com/2012/03/27/minksy-and-methodology- - wonkish/; https://krugman.blogs.nytimes.com/2012/03/27/banking mysticism/; - - https://krugman.blogs.nytimes.com/2012/03/30/banking mysticism continued/; https://krugman.blogs.nytimes.com/2012/04/01/tobin-brainard-1963/; https://krugman.blogs.nytimes.com/2012/04/02/things-i-should-not-be-wasting-time-on/; https://krugman.blogs.nytimes.com/2012/04/02/a-teachable-money-moment/; https://krugman.blogs.nytimes.com/2012/04/02/oh-my-steve-keen-edition/.

>

Figure 134: Screenshots from Krugman's promotional video

Yes it is (also) about money, as I’ll now explain using Minsky .[42] Firstly, here are the substantive parts of Krugman’s first post in 2012, where he set out very well the basic assumptions of the “Loanable Funds” model of banking. I’ve highlighted the key passages in italics:

I always try to find the simplest representation I can of whatever story I’m trying to tell about the economy. The goal, in particular, is to identify which assumptions are really crucial — and in so doing to catch yourself when you’re making implicit assumptions that can’t stand clear scrutiny.

Keen doesn’t seem to be doing that. His paper contains a number of assertions about what is crucial, without much explanation of why these things are crucial. And I guess I just don’t see it.

In particular, he asserts that putting banks in the story is essential. Now, I’m all for including the banking sector in stories where it’s relevant; but why is it so crucial to a story about debt and leverage?

Keen says that it’s because once you include banks, lending increases the money supply. OK, but why does that matter? He seems to assume that aggregate demand can’t increase unless the money supply rises, but that’s only true if the velocity of money is fixed; so have we suddenly become strict monetarists while I wasn’t looking? In the kind of model Gauti and I use, lending very much can and does increase aggregate demand, so what is the problem?

Keen then goes on to assert that lending is, by definition (at least as I understand it), an addition to aggregate demand. I guess I don’t get that at all. If I decide to cut back on my spending and stash the funds in a bank, which lends them out to someone else, this doesn’t have to represent a net increase in demand. Yes, in some (many) cases lending is associated with higher demand, because resources are being transferred to people with a higher propensity to spend; but Keen seems to be saying something else, and I’m not sure what. I think it has something to do with the notion that creating money = creating demand, but again that isn’t right in any model I understand . (Krugman 2012b. Emphasis added)

The key technical issue here is what do banks do ? According to Krugman, banks take in deposits from some customers, and lend them out to others:

If I decide to cut back on my spending and stash the funds in a bank, which lends them out to someone else, this doesn’t have to represent a net increase in demand…

This is not, of course, what banks actually do, as we now can state with the authority of the Bank of England:

This article explains how the majority of money in the modern economy is created by commercial banks making loans. Money creation in practice differs from some popular misconceptions — banks do not act simply as intermediaries, lending out deposits that savers place with them , and nor do they ‘multiply up’ central bank money to create new loans and deposits. (McLeay, Radia, and Thomas 2014. Emphasis added)

But it’s worth putting Krugman’s misconception into Minsky to show that, if Neoclassicals were right about what banks do, then they would also be right to ignore banks in their macroeconomic models.

Fundamentally, as the Bank of England notes, Neoclassicals believe that banks act “simply as intermediaries”. I call it the “Ashley Madison theory of banking”—see Figure 132 if you haven’t heard of Ashley Madison before.

Figure 135: Ashley-Madison as an intermediary

Ashley Madison doesn’t actually provide sex: instead, it lets men who want sex find women who want sex, and charges a fee for the introduction service. Similarly, in the Neoclassical mind, banks don’t actually provide money: instead, they let people with more money than they need at the moment (savers) meet people with less money than they need (borrowers). The savers lend money to the borrowers, and the bank charges a fee for the introduction. No money is created because of the new debt—just ask Paul Krugman:

Think of it this way: when debt is rising, it’s not the economy as a whole borrowing more money . It is, rather, a case of less patient people—people who for whatever reason want to spend sooner rather than later—borrowing from more patient people. (Krugman 2012a, pp. 146-147. Emphasis added)[43]

The easiest way to model what Krugman—and all Neoclassicals, with almost the sole exception of the Bank of England economist Michael Kumhof—think banks do, is to model the literal case of savers lending money directly to borrowers, through the deposit facilities provided by banks. Figure 133 shows the banking sector’s view of that person-to-person case, where the “less patient people” are factories, and the “more patient people” are rentiers who both invest in and lend to factories. The bank has only one asset—Reserves, which match the sum of the deposits of Impatient people, Patient people and Workers, plus the Banking sector’s Equity.[44] The first four rows show financial operations—lending, paying interest, repaying debt, and paying the bank’s “intermediation” fee. Then we have paying wages to workers, which enables production; dividend payments to shareholders (those “patient people” again), and finally consumption of the output of the factories managed by the “impatient people” by “patient people”, workers and bankers.

Figure 136: Patient lends to impatient via the banking system

Notice that while lending shows up in the banking sector’s Godley Table, the actual debt owed doesn’t, because in this model, the debt is not an asset of the banking sector: instead, it’s an asset for the “Patient” people and a liability for the “Impatient” people. So to see the debt itself (which I labelled as “Loans” in this model), you have to create additional tables for Patient, Impatient and Workers. Figure 134 shows all the Godley Tables in this model—as noted in the chapter on Godley

Tables, all you have to do is create a new table and then use the down-arrow on the columns ( ) to search for Liabilities that haven’t yet been defined as Assets, and vice versa.

Figure 137: All the Godley Tables for "Patient to Impatient" lending

To complete the model, I made very similar definitions to the model developed in Figure 126—see Figure 135. The main differences are that Krugman’s silly “Impatient” term takes the place of the Firm sector there.

Figure 138: The definitions of flows in the Loanable Funds model

With those definitions made, the model can be run, and the parameters that control lending and repayment varied while the model runs, to see the impact of higher and lower levels of credit and debt on this toy economy. While there clearly is some impact, some things don’t change: as Krugman put it himself, “when debt is rising, it’s not the economy as a whole borrowing more money”: changes in debt have no impact on the money supply. There are variations in GDP and incomes as a result of the variations in the time constants for lending and repayment, but they are both minor and transient. If this model described the real world accurately , then it would make sense to leave banks, debt and money out of macroeconomics. To cite Krugman once more:[45]

“I’m all for including the banking sector in stories where it’s relevant; but why is it so crucial to a story about debt and leverage?” (Krugman 2012b)

Take a minute to savor this statement. If anyone scoffs at the assertion that mainstream economists don’t understand banks, debt and money, just show them this gem.

Anyway, to answer his question, we have to take account of the real-world situation that banks lend to non-banks, so that Loans are an asset of the Banking Sector, and not of “Patient People”. Before I show how to do this, note one aspect of Figure 136: given the parameters in the model, a higher

>

debt to GDP ratio is associated with a lower GDP—see the Income and Debt/GDP plots on the right hand side of Figure 136.

Figure 139: Dramatic changes in Debt/GDP, minor transient changes in GDP

To change this model so that Banks, rather than “Patient People”, lend to “Impatient People”, you have to:

• Shift the Loans column from Patient’s Godley Table to the Bank’s;
• Delete the financial operations on the Patient Godley Table: the first three rows for Interest payments, lending and repayment all go, leaving just two rows—receiving Dividends and consuming;
• Make room for a new Asset on the Banking Sector Godley Table by clicking on the green plus icon below the Asset label;
• Click on the , which will show Loans as a Liability that hasn’t yet been classified as an Asset (when you delete the Loans column from Patient’s Godley Table, Loans remains in the model as a Liability of Impatient), and select Loans;
• Minsky then brings across the two operations that affect Loans, Lend and Repay;
• The Banking sector Godley Table will still show Interest as a transfer out of Impatient’s account, but it doesn’t go anywhere; Type “Interest” into the BankE column, to show that interest payments increase the (at-call) equity of the banking sector.

That’s it: strictly speaking we should also change how lending is determined, since the Loanable Funds model shows lending as being based on amount of money in Patient’s deposit account, but this is enough to see if this simply structural change to the model—as opposed to a behavioural change—has any impact on the dynamics.

Figure 140: Altering the Godley Tables of Loanable Funds to fit the real world

You bet it does: see Figure 138. Debt creates money, so the money supply rises when debt rises, and falls when it falls; a rising debt to GDP ratio is associated with a rising GDP (and the obverse for falling debt); credit growth, which was out of synch with GDP growth in the Loanable Funds model, now parallels—and in fact causes—the growth of GDP. We are in a completely different world to the Neoclassical model of loanable funds—and it happens to be the real world we actually inhabit. Loanable Funds is a misleading fantasy.

Figure 141: Dramatic changes in Debt/GDP, dramatic changes in GDP

So too are all the models that go with it—especially the model of “Fractional Reserve Banking”. Here Mankiw’s Macroeconomics textbook gives a good outline of the fantasy. It starts with banks as just warehouses for deposits:

We begin by imagining a world without banks. In such a world, all money takes the form of currency, and the quantity of money is simply the amount of currency that the public holds. For this discussion, suppose that there is $1,000 of currency in the economy.

Now introduce banks. At first, suppose that banks accept deposits but do not make loans. The only purpose of the banks is to provide a safe place for depositors to keep their money.

The deposits that banks have received but have not lent out are called reserves. Some reserves are held in the vaults of local banks throughout the country, but

most are held at a central bank, such as the Federal Reserve. In our hypothetical economy, all deposits are held as reserves: banks simply accept deposits, place the money in reserve, and leave the money there until the depositor makes a withdrawal or writes a check against the balance. This system is called 100– percent–reserve banking. (Mankiw 2016, p. 89. Emphasis added)

Mankiw displays this using a T-Account:

Figure 142: Mankiw's model of Full Reserve Banking

When Mankiw introduces lending, it is lending from reserves :

Now imagine that banks start to use some of their deposits to make loans… The banks must keep some reserves on hand so that reserves are available whenever depositors want to make withdrawals. But as long as the amount of new deposits approximately equals the amount of withdrawals, a bank need not keep all its deposits in reserve. Thus, bankers have an incentive to make loans. When they do so, we have fractional–reserve banking, a system under which banks keep only a fraction of their deposits in reserve. (Mankiw 2016, pp. 89-90)

Mankiw then shows the bank as lending out 80% of its reserves:

Figure 143: Mankiw's model of Fractional Reserve Lending

Firstbank’s Balance Sheet

Let’s compare this model of banks to what banks actually do, in Minsky—see Figure 141.

Figure 144: Comparing actual banking with the Fractional Reserve Banking Model

The first line shows what banks actually do to make a loan: they add an amount to the borrower’s deposit account, and simultaneously record a debt by the borrower to the bank for precisely the same amount. Lending creates deposits directly, which is creating money directly—there’s no need for the iterative process alleged in the Fractional Reserve Banking model.

The second line shows the first stage of Fractional Reserve Banking model, but it is clearly incomplete: it shows a transfer of Assets from Reserves to Loans, but where is the money for the borrower?

The only way to show the loan actually giving money to the borrower is if the loan is in cash : the borrower walks out of the bank with a debt to the bank as shown in Figure 141, and an equivalent amount of cash, as shown in Figure 142.

Figure 145: Completing the first round of Fractional Reserve Banking

This alone is enough reason to reject the model: it’s very easy to say “use some of their deposits to make loans” as Mankiw does, but when one models what that means in strict double-entry format, lending from reserves only works if all loans are in cash.[46] In the real world, almost all loans are made by crediting a deposit account.[47]

So why do Neoclassicals stick with an unrealistic and complicated model, in place of a realistic and simpler one? Because with the more complicated model, they can ignore banks and money and debt in their macroeconomics, and claim that the money supply is controlled by government policy— government reserve creation times the “money multiplier”—rather than determined by bank lending. In part, this is ideology disguised as science, but it’s also the standard reaction of a discipline to a discovery that contradicts a core belief. As the physicist who discovered quantum mechanics put it:

“a new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.” (Planck 1949, pp. 33-34)

Krugman’s reaction to the Bank of England report that rejected these textbook models is par for the course here. The Bank of England paper said nothing that hadn’t been said by many non-mainstream economists in the previous five decades (Moore 1979, 1988a; Graziani 1989; Holmes 1969), but it said it with the authority of a body that Neoclassical economists could not ignore:

The reality of how money is created today differs from the description found in some economics textbooks :

• Rather than banks receiving deposits when households save and then lending them out, bank lending creates deposits .
• In normal times, the central bank does not fix the amount of money in circulation, nor is central bank money ‘multiplied up’ into more loans and deposits. (McLeay et al. 2014a, p. 1. Emphasis added)

Pretty definitive, right? So how did Krugman react to it?[48] See Figure 143:

Figure 146: Krugman's reaction to the Bank of England paper

That reaction can be summarized as “I’ve read it. So what?”. He did not even consider that he should model what this meant for money creation, let alone macroeconomics as a whole. The same applies to Mankiw, whose textbook post-dates the Bank of England paper, and yet repeats the myths that the Bank of England debunked.

In a moment, we’ll leave these barter mystics behind, and consider the actual macroeconomics of money. But beforehand, I want to cover one important point: though all the Minsky models I’ve shown to date have been either pure “Godley Table” models, or pure flowchart models, it’s quite easy to mix the two—thus letting Godley Tables handle the monetary dynamics of a model economy, and the flowchart cover the physical dynamics.

7.6 A Mixed Godley Table-Flowchart model

In all the models to date, I’ve either used the Godley Tables to lay out the system’s dynamics, as in Figure 134, or a flowchart, as in Figure 104. In every Figure between Figure 121 and Figure 138, the

>

flowchart components were used to define variables in the Godley Table itself, while the Godley Table generated all the differential equations that power the model.

But this is done just for convenience. It’s quite easy to combine a model of mixed monetary and physical dynamics. I’ll illustrate this starting from a simple Godley-Table-only model, as illustrated in Figure 144.

I’ve copped a fair bit of criticism (on YouTube and social media) for the part of this model highlighted in Figure 144: the determination of GDP by turnover of money in the Firm sector’s bank account, and then also profits and wages.[49]

Figure 147: A pure Godley Table model with money turnover assumptions determining GDP, Profits and Wages

For the record, I don’t think this is how capitalism actually works: this is just a genuine simplifying assumption to allow me to ignore the physical economy and concentrate on monetary dynamics, which—thanks to the ignorance of Neoclassical economists and the fervency of other ideologues— remains an area of great contention in economics today. Of course, Neoclassical economists get the physical economy wrong as well, with their fetish for equilibrium, and an unrealistic “production function” (the “Cobb-Douglas”) that is both tautological where it’s right, and delusional where it’s wrong.[50] In practice, I want to encourage economists to build mixed monetary-physical economy models (with ecological linkages as well—see Chapter 10, starting on page 195). So let’s see how, using my model of Minsky’s Financial Instability Hypothesis.

Figure 145 shows this model. The physical output part of the model—the flowchart components directly below the Godley Table—reproduce Goodwin’s model. The monetary components— borrowing and debt repayment, interest, wages and consumption—are added via the Godley Table for the banking sector. The key factors I added to Goodwin’s model to reproduce Minsky’s Hypothesis were an investment function in place of Goodwin’s assumption that all profits were invested, and debt financing investment in excess of profits. This was easily modelled by the

And some of the flowchart elements highlighted in the Figure which equate borrowing to gross investment and profits to debt repayment:

Done in this way, the crucial role of the endogeneity of money in Minsky’s Hypothesis is obvious: since loans create deposits, the act of borrowing to finance investment expands both the money supply and aggregate demand.

Figure 148: Using a Godley Table in the Keen-Minsky model

I’ve added one tiny feature to the model as well, using a switch to make it an option in running the model. In the original model, the role of consumption is effectively ignored, with workers’ consumption equalling wages and bank spending equalling interest. That is the default shown in Figure 145, but if the parameter Is set to 1, consumption is instead based on the money in Workers and Banks, divided by a time-constant: τ?

This is to address one criticism I’ve also heard of the model, and which is surely reproduced in some academic papers somewhere, that it is a “supply driven” model, rather than a demand driven one.

Hello ? The driving force in the model is the investment function, and since when was investment not a component of aggregate demand? The element that is missing from the original model (Keen 1995) was any model for consumption, so that consumption was the residual variable—since the model determined both investment and output.

The capacity to base consumption on a time constant in this model is a simple step towards modelling both investment and consumption. This necessarily means that the residual variable is now something not modelled here: unsold stocks, since with investment, consumption and output determined, the free variable will be unsold stocks. Price dynamics are also absent—though they can easily be added.

Now let’s turn to the key issue for which Minsky was designed: to allow the easy analysis of monetary dynamics in a capitalist economy.

Footnotes

34 Shortly after this manual is published, a third method of being able to add stocks and flows to the canvas from a Godley Table will exist. We plan to generalize the current system of inserting variables, parameters and constants to having a drop-down menu for each starting with “New…”, and the listing the existing stocks, flows and parameters respectively for selection to place onto the canvas.

35 https://www.marxists.org/archive/marx/works/1885-c2/ch07.htm

36 A more complex model can allow for a long-lasting pandemic and make variable. This is one of the beauties of system dynamics models compared to Neoclassical modelling with its fetish for pretending that everything happens in equilibrium. In a system dynamics model, simplifying assumptions can be dropped and a N more complex model developed, without having to throw in additional ridiculous assumptions to maintain the fiction of equilibrium.

37 No relation.

38 My birthday is March 28th, and since I lived in Sydney then, I first saw his column when I was alerted to it on my birthday by a Facebook message.

40 My posts on Krugman http://www.debtdeflation.com/blogs/2012/03/29/krugman-on-or-maybe-off-keen/; http://www.debtdeflation.com/blogs/2012/04/02/blog-observations-on-krugman/; http://www.debtdeflation.com/blogs/2012/04/03/oh-my-paul-krugman/; http://www.debtdeflation.com/blogs/2012/04/04/krugman-apologises/; http://www.debtdeflation.com/blogs/2012/04/09/capital-account-interview-on-the-keen-krugman-brawl/.

41 https://www.masterclass.com/classes/paul-krugman-teaches-economics-and-society/. Of course, this is a course that I don’t recommend you undertaking!

42 I received the INET grant that enabled me to create Minsky in September 2011, so Minsky was in its infancy back then, and in particular, we hadn’t yet implemented Godley Tables: all Minsky could do back then was model simple ordinary differential equations using the flowchart paradigm of conventional system dynamics programs—See https://www.ineteconomics.org/research/grants/extending-macroeconomics-and-developinga-dynamic-monetary-simulation-tool Therefore, I couldn’t use Minsky to illustrate my argument back then (the original version is still accessible from https://sourceforge.net/projects/minsky/files/Windows%20Binaries/).

43 I will never cease to be amused by Neoclassical protestations that their approach is “value-free”, while at the same time they use pejorative terms all the time: “perfect” competition, Pareto “optimal”, etc. And here, Krugman gives us “patient” versus “impatient” people…

44 In most models, for the same of simplicity, I treat Bank Equity as short-term “at call” funds, ignoring that banks have long-term equity (which includes long-term debt). But Minsky supports multiple Equity columns, so if you wish you can model Bank Equity as including both at-call and long-term components: just choose “Enable multiple Equity columns” from the Preferences form on the Options menu.

45 https://krugman.blogs.nytimes.com/2012/03/27/minksy-and-methodology-wonkish/

46 Or some other negotiable instrument like a bank cheque.

47 It doesn’t have to credit the depositor’s account per se : if you use your credit card to shop, the deposit account that will be credited will be the merchant’s, while the increased debt will be recorded against your credit card account. But the essence of the entire process is contained in that one line in a Godley Table.

48 https://krugman.blogs.nytimes.com/2014/04/28/a-monetary-puzzle/ .

49 Capitalist consumption and investment is necessarily implicit in this model, because expenditure by capitalists on themselves—firms buying goods off other firms, capitalists (who are subsumed into the firm sector) buying goods off firms, etc—can’t be displayed without adding several more sectoral accounts. The other option—which is a design ambition for Minsky , and is therefore dependent on raising more development funds—is to enable Godley Tables to be three-dimensional. Then intra-firm monetary exchanges would occur between slices of a Godley Cube, while inter-sectoral exchanges would occur on the front face of the cube, as now.

50 See Forget the “Cobb-Douglas Production Function” (an optional read), starting on page 193.