I’m not happy with this post. I tried to mix some rather speculative economic thinking with an attempt to explain to a wide audience, and it doesn’t work. I’ll rewrite it as geeky economic article.
The asset bubble that started in the late 1990s and exploded in 2007 as the financial crisis was caused, in my opinion, by our monetary system. In particular, the following cycle took place:
- The general public in western, mainly Anglo-Saxon, economies started using real estate as hard money, profiting from its parasitic appreciation linked to GDP growth. The real economy deflated against housing.
- Banks issued new money backed by the rising real estate. This broke monetary policy by expanding the money supply first as intended but then beyond, as banks used securitized debt to evade regulation and recycle their license to create money and use it as their capital.
- A positive feedback loop developed, where appreciating houses led to banks issuing more money, which led to inflation of money against housing. The market responded by raising house prices further, until both housing and housing-backed money crashed.
The system of money used by western economies, although no secret, is not widely understood by the public. I’ll explain how our monetary system works, how it caused the crisis, and how it ought to be reformed in principle. Obviously I have no tried and tested new system to propose, but I’ll try to articulate what new conditions it should meet.
Hard money and its parasitic appreciation on GDP
The traditional conception of money is as a fixed quantity, such as a ton of gold. It changes hands, and some people hoard it, but it doesn’t grow or shrink. That way the value of goods can settle against gold through the market. This “hard money” concept served well for most of history because the size of the real economy didn’t change much either. In a static economy, a gold coin buys a sack of wheat, say, now or in a hundred years. Using gold does nothing to erode inequality, but doesn’t amplify it either. Sitting on gold yields zero return, so any productive investment whose risk-adjusted real return is above zero beats that, and will probably get funded.
Since the industrial revolution, we have economies that grow rapidly in real terms. If people use gold or other hard money in such an economy a systemic effect results that amplifies inequality and denies production that ought to be funded. It works as follows: Say the real economy GDP, measuring total wheat produced etc, grows by 5% a year but the money supply is still a ton of gold. That means that sitting on a gold coin for a year buys a sack and 5% more wheat, so anyone who holds excess gold will tend to do that. The person sitting on gold didn’t do anything useful, so we can say that hard money has a parasitic appreciation linked to GDP, or that the real economy suffers deflation against the currency gold.
Deflation amplifies inequality, as anyone who hoards say 1/10th of the total gold is entitled to 1/10th of total production, which is an ever growing amount of stuff. That’s ethically bad, but not the focus of this analysis. The more practical problem is that any productive investment, say a mill, needs to yield more than 5% risk-adjusted return, or above GDP growth, to beat the parasitic gain and get funded. Investments can’t beat GDP growth on average, by definition. Something like 5% GDP growth is the aggregate result of many investments with different individual returns such as 1%, 6%, -3%, 5%, 10%, -7%, 2% etc. All those with positive return ought to get funded, but capital is motivated to fund only those above 5% return and pull out of the others. That results in violent fluctuations of real economic output and can stall overall growth.
Western economies eventually figured out this problem and abandoned gold as a result of the Great Depression of the 1930s. But then they needed a new currency to replace gold. The abstract properties of a currency for a growing economy ought to be that the money supply grows to match the real economy in aggregate, and preferably that the influx of currency gets to the hands of the people who make the growth happen (growing industries, young people entering the workforce, etc) to reward them. Unfortunately, there is no handy metal or other physical commodity with these properties, so artificial currencies were set up.
The role of banks in the monetary system
The money we use today looks to the consumer like state-issued currency but it is in fact a hybrid. Roughly 10% is issued by states (central banks) and 90% is issued by private banks. The kind that’s printed on colored paper and says €20, or $20, or whatever is issued by states. It’s printed by central banks, which states can in theory put under democratic control. Increasingly they don’t – they allow the central banks, notably the ECB, to serve the interests of business rather than the public, and again that’s bad but not central to this analysis. The independent central banks and their allegiance to business is a contributing factor, not the cause, of the crisis.
The depth of the problem is with the 90% or so of money that’s issued by private banks. That’s your average high street banks. They issue 90% or more of money, even if they don’t have some archaic right to print paper banknotes (bills) like they do in a few places. That bulk of the money supply is not paper money. It’s the numbers that you, other people, and regular businesses see in your bank account. The money in your bank account, even though nominally euros, dollars, etc, is not issued by the state and stored by the bank. It’s money issued by the bank. Strictly speaking it’s not issued by that specific bank you’re the customer of, but it’s issued by all the private banks working as a system.
The process is as follows. Say you want to buy a house that’s newly built and worth €100k. You go to the bank and say “Hey, can I have €100k? I’ll take a mortgage on that house”. The bank says “Sure, there you are, you have €100k. You owe us €100k plus interest, and if you don’t pay we’ll take the house”. At this point all that the bank has done is write on one side of its books that it promises to pay you €100k whenever you ask. That’s money freshly created by the bank, and you see it as a positive number in your account. On the other side of its books the bank writes it has an asset: Either regular payments or €100k plus interest from you over many years, or a house worth €100k. The bank is allowed to create money in this way because the entries are balanced. The sum is zero or in the bank’s favor, so the bank will in theory be able to make its promises to pay. That one bank will actually have to pay you the €100k when you withdraw it to give it to the house builders but they’ll immediately put it in a bank account in the same or another bank. The transfers of money between banks balance out so that banks only need a small amount of paper money, issued by the government, to cover cash withdrawals or other transactions that leave the banking system. Thus banks create and then maintain the bulk of money as nothing but entries in their books.
Banks are granted special legal rights, not available to regular businesses or individuals, so that they can implement this privately run monetary system. So long as it meets certain accounting metrics, a bank is allowed to issue money by simply declaring a promise to pay someone. It can’t print it on paper notes payable to “the bearer”, but it can hold it as an account payable to a specific person. Credit in the account is legal tender because the law treats IOUs from banks as money. It’s money in substance because the bank can pressure the people who owe it loans to pay their dues, or take their property and sell it. It’s practical money because the bank nowadays provides a fast electronic payment system that moves credit between different people’s accounts across the banking system. So the money that we use now is roughly 10% old-fashioned state currency and 90% or more privately issued credit backed by the general public’s debt obligations or property in mortgage.
The reason for the 10%/90% hybrid is to allow the state to regulate monetary policy. Taking the US as example, there is roughly $10 trillion in total, but the Fed only issues about $1 trillion. US law allows banks to store a dollar of Fed-issued currency as reserve and issue ten dollars of their own money, so that the amount issued by the Fed and that 10x legal multiplied determines the total amount of dollars. Say the Fed measures that US GDP has grown by 2%. That means it must increase the overall supply of dollars by 2% to avoid deflation, so it has to print an extra $20 billion. The Fed lends this currency to banks, which store it and are therefore allowed to issue an extra $200 billion of their own money. The money supply grows by 2% as intended by policy, while it’s up to the banks to find customers with credible assets or income and convince them to take loans to back the $200 billion that they issue.
This system is an ingenious attempt to avoid the problem of deflation caused by using gold, or other hard money in a growing real economy. Mostly it works, and there is no superior system widely proposed. That said, let us enumerate its shortcomings:
- The system motivates banks to enlarge the money supply as much as possible since more loans directly relate to more intrest revenue. There is a conflict between banks’ profits and the monetary policy they are licensed to implement, and that is the main problem that led to the present crisis.
- Banks should ideally give the new money to people and businesses who promote real GDP growth, such as investing in R&D or job creation. It doesn’t fulfil policy goals if the new money joins existing idle capital. That is a major contributor to the crisis.
- The system grants a license to banks, as providers of the monetary system, to obtain essentially free capital which they convert to loans and charge interest on. That is a public asset that enriches the banks substantially, perhaps more than any of their other business assets and services. Arguably it has not been granted to banks transparently by the public will, and the public is not being adequately compensated.
- Because there is no separation of the bank as an operating business and the monetary fund that it runs, the economy is hostage to the banks. Any political will to hold the banks accountable, or any fanciful idea to blow them up, cannot proceed because if anything bad happens to the banks the money that they issue as IOUs disappears. That is archaic and should be straightforwardly fixed by separating the bank’s operating balance sheet (its offices, staff costs, etc) from its trust balance sheet (the loans and deposits that it runs). The proponents of narrow banking advocate this change, and I agree, even though neither of us I think wants to blow up banks.
- For the same reason, even if anything accidental happens to the banks they can pass the harm onto the public, which forces government to provide free unlimited insurance for their activities. This is the part of a crisis that most enrages the public, although it’s not the most harmful part. Even in a healthy monetary and banking system, there needs to be deposit insurance. Banks have to pay for insurance against their individual failures, as some fairly mainstream reformers of the system argue. The monetary system needs to be resilient, rather than insured, against their systemic failures.
Let’s look now at how the monetary and banking system went wrong, starting in the 1990s and into the present crisis.
Deflation of the real economy against housing
Housing can be viewed as a commodity that provides living comfort, as a capital good, or as a token asset held for its exchange value. We regard hotel accommodation as a commodity – we rent it and care about the quality of service offered. Through much of history, housing and farmland have been viewed as productive capital goods. Buildings depreciate through wear, like cars. Farms have value because they directly support production, like machinery, and some have better output than others. Feudal land holding was an especially unequal distribution of land seen as a means of production, not as a token of value. These approaches to land are economically sane.
More recently, especially in Anglo-Saxon cultures, the general public has been using housing as a token of value. House ownership makes the greater part of people’s net worth, even though their earnings are made at an office or workplace outside the house. The public delights when house prices go up, even though this makes living unaffordable. Houses are bought and sold as investments, not as devices that improve quality of life. What is being traded are licenses to live in or occupy a particular location. The building is not the important part, and often the quality and comfort of houses is sub-par where the focus on real estate as investment is strongest.
We’re using houses as large gold coins. Housing has acquired the same economic characteristics that gold had in until the 1930s:
- It is compelling. Housing is accepted universally as a token of value, and normally the market clears. People and businesses have to buy up the available urban and suburban buildings because they compete for location, and if they don’t someone will buy the titles for redevelopment. The effect is that the real economy (or the great majority of it, which is driven by city dwellers) gets priced against housing.
- It is practically in fixed supply. Although cities and suburbs can sprawl, real estate doesn’t grow or shrink significantly because the high-value part is concentrated in city centers or other small desirable locations with inflexible building rules. Since the growing economy gets priced against housing which is fixed, housing appreciates parasitically on GDP growth.
- Any particular unit retains its value indefinitely as an asset. Even if buildings crumble, titles of land holding are effectively eternal, like gold.
- Real estate, or the right to claim it as security, is held by banks as a tangible asset to back the issue of the formal currency that we use.
The result is that real estate titles start functioning like money. They take a nominal value that far exceeds their real use value, and the difference is a reflection of the value of the real economy.
Perhaps it should be no surprise that real estate is causing the same problems that gold used to. It appreciates parasitically on GDP growth. In the macro, housing appreciates because the participants in the growing real economy compete to buy up and hold the fixed supply of housing. In the micro, an appartment’s value rises because of the good work of teachers, cooks, shopkeepers, or entertainers in the city – not because of the activities of the builder or the owner. This promotes inequality as older people who have invested in housing earlier become disproportionately wealthy compared to the working young.
Because real estate appreciation will at least match GDP growth, housing beats any average or below average real investments and deprives them of the capital that ought to fund them if returns were unbiased. This in turn slows down the economy, as only the minority of high yielding investments go forward. Tech startups happen, but bakeries do not as people put their savings in houses instead. The real economy deflates, not against gold this time but against real estate.
Breaking monetary policy
As we’ve seen, the central bank regularly measures GDP growth and tries to expand the money supply by an equivalent amount, so that there is near zero inflation. Falling slightly short causes deflation, which is hard to control, so the policy aim is small positive inflation. Continuing with the US example, let’s say the money supply needs to grow by 2%. There’s roughly $10 trillion in existence, so a fresh $200 billion has to enter the economy. If every dollar were issued by the Fed, the Fed would have to print all $200 billion and hand it out essentially free to some actors in the real economy, preferably in such a way as to reward growth. It could give the money to the federal budget (the treasury) or it could have a large grant-giving agency to decide who and what to fund. Such a system might be seen in smaller economies.
Instead, the US, EU, and other large capitalist currencies have a privatized money supply. Banks maintain a balance sheet where their IOUs (current/checking accounts) are our money supply and their receivable loans, or claims to property under mortgage, are the assets that back the money. Banks are motivated to enlarge their balance sheet infinitely, since the more loans they can issue the greater their profits from interest. They’re limited by business risk, where the risk-adjusted return from loans becomes lower than the interest banks pay out to consumer accounts (or zero), and by accounting metrics imposed legally. Plenty has been written on the banks’ tendency to under-price risk, but here let’s focus on the legal limits, as these exist to constrain the banks to implement monetary policy.
Banks have to meet a solvency constraint, which means that the credible expected return from loans, or from taking and selling mortgaged houses, must be higher than the IOUs that the banks have issued as our money. This is obvious enough – otherwise banks would use their license to issue money to directly line their pockets. Instead this constraint says they’re only allowed to create money as free capital, use in sound investments, and keep the profits. Banks also have to meet a reserve constraint, which is to hold state-issued currency equal to a certain fraction of their balance sheet – usually 10%. This is the control that the central bank has to regulate how much money the banks can issue. The bank has a license to generate free capital, but only so much (ten times as much as its holding of state currency). When properly applied these controls allow bank to implement monetary policy safely and as intended, while profiting from the returns on good loans they issue.
The system goes wrong when banks subvert these measures in order to expand their balance sheet more than intended in pursuit of more profits through interest. They subvert the solvency constraint simply by lending larger sums to poorer people with less secure jobs. Lending to poor or insecure borrowers is not in itself a bad thing. It can be good if done constructively, but the risk must be correctly reflected in the value of the loan. It’s always a matter of judgment whether the borrower will repay the loan, and how much of it. By being optimistic, negligent, or duplicitous the banks have over-represented the value of their loan assets and passed the solvency bar, where objectively they would have fallen short. That much is straightforward bad practice.
More interestingly, banks subverted the reserve requirement, and hence the limit on how much free capital they can issue, through securitized debt: CDOs and other such instruments. Remember that the requirements is “you must hold $1 for every $10 on the outgoing (liabilities) side of your balance sheet”. If a bank has $10 million of federal currency it can have $100 million in liabilities such as consumer accounts, and let’s say it’s also solvent and records $130 million in receivable loans including interest. If the bank can chop off, say $20 million of its liabilities and $26 million of its receivables and sell it off as a completely unregulated free-floating entity, it’s left with $10 million of Fed currency and $80 million of its own issue. It’s then allowed to re-use its license to free capital to issue a fresh $20 million, so long as it can find consumers to sign up for loans of that amount.
A Collateralized Debt Obligation (CDO) is just that free-floating unregulated iceberg broken off from a bank’s balance sheet. You can also think of a CDO as an unregulated micro-currency. Banks have a highly regulated license to issue their home currency (dollars, euros, etc) as we’ve seen. They cannot print Citi dollars or RBS pounds outright, but with CDOs they can do something equivalent. The CDOs are identified with boring names and numbers, and they’re sold to other banks, pension funds, and other institutions rather than directly to consumers, but otherwise they’re like money. They’re traded in a liquid way, or they were until the market collapsed in 2007, they’re held as liquid assets by investors, and their value goes up and down based on broad perceptions of the market, like the value of currencies (and less like the value of shares, where the individual performance of the company matters).
From the 1990s, banks have enthusiastically used this device to break off pieces of their balance sheet, abuse their license to issue money, effectively enlarge their balance sheets beyond the permitted amount, and profit. They also in practice picked the riskiest assets to break off and sell to fools (as Stigitz describes them) so that they kept the healthier balance sheet and the fools got the risk. Again, that’s ethically bad but the main issue is that banks broke monetary policy. Even if they had sold off the most prime assets, we’d have the same problems. The damage was done over 15 years or more, as banks have been breaking their reserve constraint. It suddenly emerged as a crisis in 2007 when these free-floating assets started sinking and banks started failing their solvency constraints too.
When the crisis did break out in late 2007 the main difficulties were legal, not substantial. The rules of capitalist banking systems were designed with the assumption that a few banks might fail individually, for their own business reasons. Thus the rules require that insolvent banks be immediately liquidated, to keep the banking system healthy. The new situation was that nearly all banks were insolvent by a small percentage, so in theory the banks should have been liquidated en masse. We couldn’t do that because if we did the money that exists as IOUs from these banks would disappear, and there would be economic chaos.
Our mediaeval (literally) banking system makes money, and thus the economy, hostage to the banks. Under the circumstances, states had no choice but to keep banks afloat, and they way they chose to do it was to take on the insolvency of the banks as state insolvency. Where CDOs or other toxic assets were involved, the sate effectively issued currency to take in and cancel the failed private micro-currency of the bank. States are now dealing with this insolvency, mostly by imposing austerity to re-claim the loss from the real economy. As I argued in the previous article, a policy that reclaims the loss by eroding capital, such as inflation, would be far preferable.
Feedback, leading to asset inflation
These problems would not have got so far out of hand if each of them had acted in isolation. They are all self-limiting in some way. If house prices appreciated but the money supply would not grow as much as it did, credit would run out and house prices would stop rising. If banks attempted to abuse their money-issuing license but house prices were not rising then there would be no demand for speculative mortgage loans and banks would stop producing excess capital, or they would lend it instead to productive firms. The tragedy is that these effects have been allowed to reinforce each other, with no mechanism to detect or stop the cycle.
Recall the example where you ask the bank for €100k for a newly constructed house. You get a mortgage, which creates new money that makes its way to the house builders. That’s money supply expansion that rewards production, as per the policy intent. In two years, though, the economy has grown by 5% a year so the house is worth at least €110k. Someone else gets a mortgage and offers you €110k on the same house, of which €100k was circulated in the banking system and €10k was freshly issued. That €10k is money supply expansion funding parasitic gain to you, rather than anything productive.
More worryingly, the €10k has entered the money supply more than once. It was first issued as bank money, that is IOUs backed by consumer debt, as intended. However it was then used to buy housing, which we have seen functions as hard money. The bank has used the extra €10k of housing value either to issue fresh currency or, if the reserve limit was in the way, to issue a CDO to get round it. Other banks and institutions would have taken the CDO as an asset. We have therefore roughly three kinds of money-like assets: nominal money (dollars, euros, pounds, etc), real estate titles manifest as mortgages, and CDOs with all their variants. Nominal money is backed by the other two kinds, which is a very big problem.
It is as if, back in the 1930s, the Bank of England had issued paper pounds to reflect the market value (in pounds) of gold in its vaults but also permitted the use of gold as legal tender. If they did such a thing a positive feedback loop would quickly result in high inflation. Neither gold nor nicely printed paper have that much intrinsic value. They are money, which means they’re supposed to represent the value of the real economy. If you have several kinds of money you cannot issue one kind to represent the value of another kind unless you take the second kind out of circulation and keep it all in a vault, or otherwise make it an illiquid asset.
The modern banking system was guilty of just such an egregious double-accounting scheme. It issued money against real estate market value and real estate market value packaged up as CDOs, while cheerfully letting these assets appreciate by reflecting the growth of the real economy. The total money supply (counting money, real estate, and securitized debt) expanded at roughly twice the rate it ought to, and the banks, as well as housing speculators, profited from this inflation.
Reforming the financial system
A substantial reform of the financial system has to look back and ensure the following principles:
The banks as operating businesses need to be separated from the monetary fund that they run. There needs to be a simple operating balance sheet for the bank’s buildings, staff costs, and everything else it directly owns or owes like any other business. There needs to be a separate fund for the money that the bank issues and the loans or other receivables it holds as assets. That is one aspect of the narrow banking proposal. The point is to allow banks to fail as businesses while keeping intact the slice of the monetary system that they run. If a bank failed its operating balance sheet would be liquidated, like any other business, while the fund would be transferred to a more competent bank or caretaker organization to manage. In cases of systemic insolvency the state could bail out the fund, but not the bank, and the bank would not be able to take the bailout money to pay executive bonuses. The two are inseparable now, and by far the most urgent and effective measure is to separate them.
While we use a system of money backed by consumer and small business debt, both bank capitalization and bankruptcy need to be handled in a better way. There are already capital requirements for banks, similar to the reserve requirements we talked about. Firstly, these need to be high enough to absorb foreseeable losses. Let’s say a bank with $100 million in liabilities may be required to have $10 million of Federal Reserve currency and another $10 million of regular capital. Assuming the bank’s operations and the fund that it runs get separated, when the fund is insolvent the bank must be made to cover the loss using its capital. If the capital runs out that would generally bring down the bank as a business (unless it’s a holding company with other sizeable activities), the managers would get laid off, and investors who bought the banks shares in the stock market would lose their money as per the normal rules of capitalism. If the fund is still insolvent after the bank is liquidated, then deposit insurance would be claimed. If that is still not sufficient, because the loss is systemic or the bank in question is too big, then the losses need to be passed on to the public. With a properly separated fund, this can be done in an orderly and legally pre-arranged way. For example a state, or larger body such as the EU, can decide to pass a certain portion of the loss onto the customers of the bank (perhaps zero) and socialize the rest. Socializing the loss means printing money to cover the nominal insolvency of the money fund, so that the loss becomes inflation.
In addition to reforming the banks as such, housing has to lose its incompatible status as both hard money and an asset to back the issue of bank money. One way, favorable to the Left, would be to dampen down the real estate market significantly with taxation on transfers, nationalization, and other controls, so that housing is once again seen as commodity or a capital good and not as a store of value. This would be acceptable in some cultures much more than others. For the Western, Right-leaning economies banks can instead be regulated to disallow the use of housing as an asset to guarantee the issue of new money. Banks would be able to generate money as before and issue loans for any kind of productive activity, including construction or renting, but not for the purchase of real estate. They would have to raise regular capital from the investment market if they want to sell mortgages. The effect would be to greatly reduce the amount of capital available to buy housing and to cut the cycle of housing appreciation and inflation.
The legal license of banks to issue their own capital, embodied in the 10% reserve requirement, needs to be recognized as a public asset of very high value and rented, not given freely, to the banks. The narrow banking people would withdraw it altogether, but I have doubts whether the market would provide adequate and cheap capital in that case. In my opinion the facility needs to be properly priced, so that some of the benefit accrues to the state and can be used to fund deposit insurance or other public goods. The license of banks to issue capital also has to be properly policed. If banks once again use instruments like CDOs to break off pieces of their balance scheet they would have to pay back the corresponding reserve, so that CDOs can no longer be used to evade monetary policy (they do have other legitimate uses).
In the long run, society has to consider whether a monetary system based on consumer debt, mediated through private lenders, is a good idea. On the plus side it’s a de-centralized system where many banks, rather than a central agency, have to find customers for loans. As a market system it has a better chance to cope with complexity and identify productive recipients for investment. Like any capitalist system, it pays a known and somewhat controlled fraction of its turnover to self-interest, as profit, rather than losing an unknown and uncontrolled fraction of its turnover to self interest through corruption.
Other than that, in general, debt is bad. There are only two valid reasons to take it. One is to recover from short-term financial stress. The other is to pursue a productive investment which you could not pursue simply by saving. If the financial system were perfectly efficient, meaning very low capital costs, negligible profit for banks, and no asset appreciation other than through the direct productivity of said assets, then a monetary system like ours would be a fantastic idea. Money would be issued in line with the expansion of the real economy, and the new money would go to those who make the world productive.
We’re very far from this ideal. Banks are monopolistic rent-seeking organizations rather than efficient providers of an infrastructure for allocating capital. The cost of capital is too high, partly because of bank greed and partly because asset appreciation is allowed to compete with real returns. The result is a financial system that only works well, if at all, for short-term debt. It’s neither efficient nor structured correctly for finance that extends over decades. Thus debt, the way it exists today, is not a suitable commodity to back our monetary system.
A new monetary system that ties the issue of money to real production must be found. I don’t know what that would look like. Indeed inventing it would be a momentous task. Most likely, monetary expansion needs to be coupled to human life. As an idealistic example example, every person could issue a “note of gratitude” (NOG) per day in their life and use it to pay for goods and services. Money expansion would thus be egalitarian, and people could still get rich by having a successful business that accumulates NOGs, pays them as dividends, etc. Old NOGs should relatively depreciate, which agrees with our intuitive concept of gratitude and allows the money supply to be tuned to size. Such as system would be computationally complex, but probably within reach.
A more pragmatic idea might be to adapt existing systems such that monetary expansion is, as far as possible, permitted only for real growth and disallowed for asset appreciation. If someone builds a house or a bakery, this is value creation and can be financed with newly printed money. If the same building appreciates next year, without actually improving, it’s asset appreciation. It should be legal to finance asset appreciation only by transfer of money, in a zero-sum way, so that some assets can appreciate in relation to other assets to reflect their true desirability. It should be illegal for banks to use their money creation abilities to finance the appreciation of assets in aggregate.
I wonder if it might be possible to connect money creation to value-add creation. Assuming we have a means to measure value-add, as distinct from asset appreciation, then perhaps the money supply could be expanded through VAT (value-add tax) rebates. When a chain of production takes place, businesses generally transfer value tax-free down the chain and the eventual consumers pay the value-add tax for the sum of everyone’s value increments. If this is a fairly accurate and equitable measure of value creation, the monetary authority can expand the money supply by rebating the appropriate fraction of value-add, either to the consumers or to the producers.
This article was composed over several years, with many false starts, and much of my reading material was from news and other ephemeral sources that I did not record as references. However a few sources stand out:
The intermediate-level textbook The Economics of Money Banking and Financial Markets gives a quality, if lengthy, explanation of how the banking system actually works.
The Wikipedia entries on the money supply and fractional reserve banking are very well written. These and related articles give a good working explanation of the concept of money creation by banks. Money issued by the state is called M0, and money issued by banks is called M1, M2, etc. In this article I’m complaining about M1, M2, etc.
Young, British, very left-liberal economist and journalist Philippe Legrain gives a refreshing and uncompromising account of the crisis and looks deeper into the systemic problems of the economy that led to it, like I try to do here. Listen to his LSE talk (card, audio, video), or read his book Aftershock.
US economist Laurence Kotlikoff is a very articulate proponent of narrow banking. I agree with some of his key ideas, although not with his overall stance on economics or politics, as far as I can determine. He has a blog, or you can listen to his LSE lecture (card, audio) or buy his book oddly titled Jimmy Stewart is dead.
Finance geeks may take an interest in the website of the Bank of International Settlements, a bank where central banks bank. It contains statistics about the financial positions of the banking systems of different countries, as well as some policy statements.
While not directly related to this topic, the forum VoxEU.org has many scholarly but accessible and politically interesting articles on modern economic issues. It is mostly left leaning, but also hosts some liberal or right-leaning articles.
The forum Project-Syndicate.org also makes interesting reading, though it’s mostly center-right and editorial rather than academic.
Finally the NPR programme Planet Money introduces many issues about the workings of the economy, and of course the crisis, in a rigorous but surprisingly fun and accessible manner.