Notes on ‘Notes on the Trade cycle’
Chapter 22 and fiscal monetarism, a deliberate contradiction in terms.
Stephen John Richmond
This essay is an analysis of Chapter 22 of John Maynard Keynes’ ‘General Theory of Employment, Interest and Money’- ‘Notes on the trade Cycle’ in the context of subsequent economic theories and analysis such as Monetarism and Modern Monetary Theory. It argues that by combining some Keynesian ideas with later Monetarist ideas while using a Georgist central bank (collecting taxes on rentier income and paying out a citizen’s dividend) it is possible to satisfy the concerns Keynes’ laid out in General Theory without running afoul of ever higher inflation.
- Analysis of Chapter 22: ‘Notes on the Trade Cycle’
- The Phillips curve and Monetarist critiques of Keynesianism in the context of Chapter 22
- Low unemployment with stagnant wages
- Modern Monetary Theory and full employment of labour in the context of Chapter 22
- A few notes on Neo-Fisherism
- Fiscal Monetarism: a deliberate contradiction in terms
Analysis of Chapter 22:
Chapter 22 of Keynes’ General Theory describes his view of the economic cycle and what actions should be taken to moderate it. It is therefore a vital chapter in understanding the policy implications of Keynesianism. The thoughts expressed in chapter 22 are interesting when evaluating our current fiscal and monetary policy framework and what, if any, changes should be made after 2008 and the subsequent sluggish economic recovery.
Keynes opens by describing what a trade cycle is and arguing that, in his view, the trade cycle hinges on liquidity preference, marginal efficiency of capital and expected returns for investors. His view is that crisis takes hold when investors have overestimated future yield on assets and so, while the asset itself may be profitable from a societal point of view, the drop in the value in that asset due to it’s lower than expected yield leaves the investor panicked and at a loss:
“The later stages of the boom are characterised by optimistic expectations as to the future yields of capital goods sufficiently strong to offset their growing abundance and the rising costs of production and, probably, a rise in the rate of interest also. It is of the nature of organised investment markets, under the influence of purchasers largely ignorant of what they are buying and of speculators who are more concerned with forecasting the next shift of market sentiment than with a reasonable estimate of the future yield of capital assets, that, when dissolution falls upon an over-optimistic and over-bought market, it should fall with sudden and even catastrophic force. Moreover, the dismay and uncertainty as to the future which accompanies a collapse in the marginal efficiency of capital naturally precipitates a sharp rise in liquidity-preference - and hence a rise in the rate of interest.”
It is important to note here that Keynes isn’t arguing that the problem is caused by rising interest rates, as such, but instead that the sudden realisation that future yields will not match expected returns causes a rush to liquidity as investors want safe and certain assets, like cash or government bonds, over marketplace investments. What he is describing, effectively, is investors buying assets that are unable to yield the expected returns followed by, and resulting in, a credit crunch. It’s hard not to see this as a fairly on the nose description of the 2008 crash where investors had been purchasing CDOs, and bank stocks of banks that themselves owned CDOs, that they didn’t fully understand and then once they realised that the yield would be far below expectations, due to defaults and the housing market crash, the rush to safe assets (i.e. a higher liquidity preference) resulted in a credit crunch. The solution provided by monetarists is for the central bank to simply lower the rate of interest until the mismatch is corrected. Keynes argued this wouldn’t be effective:
“Later on a decline in the rate of interest may be a great help in the recovery and, probably, a necessary condition of it. But, for the moment, the collapse in marginal efficiency of capital may be so complete that no practicable reduction in the rate of interest will be enough. If a reduction in the rate of interest was capable of providing a remedy by itself, it might be possible to achieve a recovery without the elapse of any considerable interval of time and by means more or less directly under the control of the monetary authority. But, in fact, this is not usually the case; and it is not so easy to revive the marginal efficiency of capital determined, as it is, by the uncontrollable and disobedient psychology of the business world. It is the return to confidence, to speak in ordinary language, which is so insusceptible to control in individualistic capitalism.”
Keynes is clearly arguing against what he, later in the same paragraph, calls a “purely monetary” remedy. Given that we now know that society eventually transitioned to a Monetarist philosophy based on “purely monetary” remedies this is a pretty interesting statement. Keynes goes on to provide further critique of purely monetary remedies, which will be discussed in this section, but monetary critiques of Keynes will be discussed in a later section of this essay before an attempt to amalgamate the two positions in a way that solves both sets of criticisms. Also it’s worth acknowledging that while Monetarist policy did seem to be effective at bringing down inflation and thus solving the problem it had with Keynes, it did then fail to function as a cure for the great recession much as Keynes predicted. Monetarism has not failed in every recession everywhere though and it now seems clear that some recessions are shallow enough, and the scope of monetary flexibility wide enough, that interest rates alone can be used. Australia is a good example of this. What this seems to imply though is not that Keynes is outright wrong in his analysis but that circumstances vary to an extent where there are exceptions to his rule. If an economy produces raw materials with consistent demand from an external actor then the problems Keynes describes are largely circumvented. That’s what happened in Australia due to its mining industry and demand from China.
Keynes makes several other key points:
“Unfortunately a serious fall in the marginal efficiency of capital also tends to affect adversely the propensity to consume.”
“In conditions of laissez-faire the avoidance of large fluctuations in employment may, therefore, prove impossible without a far reaching change in the psychology of investment markets such as there is no reason to expect. I conclude that the duty of ordering the current volume of investment cannot safely be left in private hands.”
This is, of course, the core of Keynesian thought- during a slump net government spending must be used (either through additional spending or maintaining current spending while simultaneously enacting tax cuts) to stabilize aggregate demand. Before moving onto the Monetarist critique of this position it is important to discuss the other side of interest rate policy- raising interest rates to avoid a bubble. The argument against assuming that lower rates of interest during a recession may be insufficient to end that recession is that lowering interest rates doesn’t solve aggregate demand issues, the same argument isn’t applicable to the raising of rates to pop a bubble though and to argue against that Keynes uses a different argument:
“There is, indeed, force in the argument that a high rate of interest is much more effective against a boom than a low rate of interest against a slump. To infer these conclusions from the above, however, misinterprets my analysis; and would, according to my way of thinking, involve serious error. For the term ‘over-investment’ is ambiguous. It may refer to investments that are destined to disappoint the expectations which prompted them or for which there is no use in conditions of severe unemployment, or it may indicate a state of affairs where every kind of capital goods is so abundant that there is no new investment to be expected, even in conditions of full employment, to earn in the course of its life more than it’s replacement cost. It is only the latter state of affairs which is one of over investment, strictly speaking, in a sense that any further investment would be a sheer waste of resources. Moreover, even if over investment in this sense was a normal characteristic of the boom, the remedy would not lie in clapping on a high rate of interest which would probably deter some useful investments and might further diminish the propensity to consume, but in taking drastic steps, by redistributing incomes or otherwise, to stimulate the propensity to consume.”
Or put another way:
“The situation, which I am indicating is typical, is not one in which capital is so abundant that the community as whole has no reasonable use for any more, but where investment is being made in conditions which are unstable and cannot endure, because it is prompted by expectations that are destined to disappoint.”
By this analysis raising interest rates may ultimately pop the boom but it doesn’t address the underlying issue. Indeed rising interests rates are arguably a symptom of the boom itself. People’s over exuberance leads to an expectation of unsustainably high returns and using monetary policy to push up yields on cash merely leans into this process until firms and individuals can not afford to accept the investment at such high interest. The bubble is popped by the central bank trying to have even more over-exuberance than the unrealistically high expectations of private investors themselves!
“Thus an increase in the rate of interest, as a remedy for the state of affairs arising out of a prolonged period of abnormally heavy new investment, belongs to the species of remedy which cures the disease by killing the patient.”
Keynes therefore argues:
“Thus the remedy for the boom is not a higher rate of interest but a lower rate of interest! For that may enable the so called boom to last. The right remedy for the trade cycle is not to be found in abolishing booms and thus keeping us in a state of semi-slump but abolishing slumps and keeping us in a quasi-boom.”
This low rate of interest argument runs parallel to a second policy position Keynes espouses in this chapter:
“The remedy would lie in various measures designed to increase the propensity to consume by the redistribution of incomes or other-wise; so that a given level of employment would require a smaller level of current investment to support it.”
These points together are interesting because it appears Keynes is arguing for a low and presumably relatively consistent rate of interest with redistribution used to stabilize aggregate demand during a slump. This isn’t quite the same as the post war consensus Phillips curve Keynesianism where government purchases were fluctuated with the business cycle in an attempt to stabilize aggregate demand. I’m not arguing that Keynes didn’t argue for this at many points but it is worth pointing out the different approaches here. While the post war consensus relies on indirectly ramping up aggregate demand through government purchases it is just as possible to use redistribution for the same end. It will be the argument of this essay that it is this avenue of policy that has been insufficiently examined by either Keynesian or Monetarist economists.
Keynes appears to preemptively address monetarism further in the rest of the chapter.
“Another school of thought finds the solution to the trade cycle, not in increasing either consumption or investment, but in diminishing the supply of labour seeking employment; i.e. by redistributing the existing volume of employment without increasing employment or output. This seems to me a premature policy - much more clearly so than the plan of increasing consumption. A point comes where every individual weighs the advantages of increased leisure against increased income. But at present the evidence is, I think, strong that the majority of individuals would prefer increased income to increased leisure; and I see no sufficient reason for compelling those who would prefer more income to enjoy more leisure.”
At first glance this doesn’t necessarily seem related to modern Monetarism, which surely hopes to spur investment with lower interest rates, but it is more relevant than it first seems. Modern Monetarism is not merely about controlling the trade cycle but also about controlling inflation. Indeed it has its origin in the inflation crisis of the 1970s. If we are to understand how to address both the concerns Keynes had with “purely monetary policies” and also Monetarist concerns with Keynes, then this is the vital point of conflict. Modern Monetary theory (which draws very heavily from Keynes and this passage in particular) describes Monetarist policy as disciplining inflation by using a “buffer stock of unemployment”. This is a description of the NAIRU concept with helps connect modern central banking practices through to the above quote from Keynes. The argument goes that full employment policy created via government purchases during a slump creates a cycle of ever higher inflation due to workers having, de facto, infinite negotiating power for wages. The continuous push to lower unemployment below the “natural rate” that the economy can naturally bear is unsustainable and therefore we must use monetary policy to interrupt the business cycle, not necessarily based on avoiding a boom per se, but instead whenever workers are gaining too much negotiating power and therefore pushing wages which, in turn, pushes up inflation. Therefore workers must be made to worry, at least somewhat, about unemployment or else wages will cycle upwards. This is, in effect, disciplining inflation using a buffer stock of unemployment. Keynes continues with a passage that seems to describe this further:
“It may appear extraordinary that any school of thought should exist which finds the solution to the trade cycle in checking the boom in its early stages by a higher rate of interest. The only line of argument, along which any justification can be considered, is that put forward by Mr D H Robertson, who assumes, in effect, that full unemployment is an impractical ideal and that the best that we can hope for is a level of employment much more stable than at present and averaging, perhaps, a little higher.”
This is more or less what monetarists are arguing in the short run due to inflation concerns. The only hope, as they see it, is to lower the unemployment rate required to curb inflation over the long run.
The Phillips curve and Monetary criticisms of Keynesianism in the context of Chapter 22:
This brings us to the 1970s and concerns about how Keynesianism’s impact on inflation was unsustainable. The pre-1970s theory of inflation was the Phillips curve which showed that as unemployment got lower inflation would gradually trend higher. This chimes with Keynes’ general view of inflation in General Theory which broadly advocates that although other scarce resources may get used up, inflation is only really a problem once labour resources are used up, because as labour is added this creates additional supply of goods and services which soaks up inflation.
With spiraling inflation in the 1970s the Phillips curve came into doubt and Monetarism proposed a new theory that, although the Phillips curve may possibly be correct in the short term, in the long term the economy would become overextended and be unable to provide enough jobs to keep unemployment below it’s long run level. As the economy trips over it’s feet trying to keep up, the government pumps demand into the economy and workers gain de facto infinite negotiating power for wages. Inflation will then cycle upwards and therefore so will prices.
The Monetarist solution is to limit the supply of money via interest rates in order to discipline inflation. In practice, therefore, this puts Monetarism in fairly direct conflict with Chapter 22. However what I wish to show here is that while the Monetarist view of inflation does clash with Keynes’ view of inflation expressed elsewhere in General Theory it’s criticisms of post-war consensus Keynesianism on the topic of inflation do not necessarily have to clash with some of the ideas in Chapter 22. (Even though Chapter 22 itself clearly disagrees with Monetarism on other issues as laid out in the previous section!)
The Milton Friedman view was that this inflation was entirely caused by the monetary authority facilitating the policy of full employment and therefore increasing the money supply to match. He based this on research he had done on previous inflation crises which had all involved increases in the money supply. Velocity of money was therefore unimportant. This evidence doesn’t seem to quite imply that velocity of money never changes or can never be important though, even if it were to turn out that it rarely is. The NAIRU critique of post war Keynesianism would still work, at least to some extent, with a fixed money supply.
If there is a natural level of unemployment and the monetary system of an economy is a fixed money supply with no fractional reserve lending then the government can still borrow money to stimulate full employment. If a natural level of unemployment does exist then the economy will again be unable to keep this up, workers bargaining power will get too high and wages will rise. As firms would be unable to produce more goods and services prices will therefore have to rise to meet the labour costs. The government will have to borrow more and more money to achieve this and presumably two things would happen:
Investors appetite for safe assets would be used up at the current interest rate.
Investors will start to worry that the government (which in this scenario isn’t printing any money) will default and so the government debt isn’t a safe asset anyway!
Both of these result in higher interest rates which should pop the bubble and the government should find the full employment policy impossible to sustain. Here it was velocity of money that allowed inflation, not a change in the money supply. The difference though was that the fixed supply of money ultimately made it impossible for the government to continue the policy. Regardless this is fairly clearly not how our modern monetary system works.
In the real world we do have a central bank that issues currency and we do have fractional lending. Given that this is the case, altering the day to day interest rate will alter how much money is lent out and to whom and therefore a higher interest rate in our monetary system can have the same net effect as the high interest rate in the fixed supply of money example- by raising interest rates it is possible to choke off the economy until inflation stabilizes at a low level. This is because although the money supply isn’t fixed, the central bank is limiting it so that interest rates rise to choke off the boom enough to stop inflation.
What is important is that this does not actually counteract the complaints Keynes makes in Chapter 22 about “purely monetary” policy. Chapter 22 meanwhile does nothing to address Monetarist concerns about inflation under post war Keynesianism. They are mutually compatible criticisms. It is possible to simultaneously believe that post war Keynesianism pushed the economy beyond its limits, while giving workers unchecked negotiating power for wages AND believe that purely monetary policies don’t work well (or sometimes at all) during recessions and that killing off a boom via higher interest rates is really just the central bank copying the over exuberant expectations of returns on debt as investors, before dropping it’s expectations wildly once the bubble is popped.
Low unemployment with stagnant wages:
Before discussing how to reconcile these issues in such a way as to address both sets of criticisms it’s important to talk about long term changes to the natural level of unemployment.
A topic of much debate and consternation in the US in the spring of 2018 was why unemployment was so low but wages weren’t rising. (The median wage adjusted for inflation actually fell a little.) With unemployment below 4% the US economy was at, in theory, “full employment”. Full employment under the monetarist framework is less about everyone who wants a job having one and more about unemployment not being so low as to cause high inflation. Inflation had gradually been going up (although slowly) but wages hadn’t kept up.
As a reason I would suggest the fact that as productivity and GDP have risen considerably over the last few decades but wages have stayed almost entirely stagnant once you adjust for inflation. I.e. wages now take home a smaller and smaller percentage of an economic pie that is getting larger and larger. So the median wage has about the same buying power even though the economy has grown enormously. This has coincided with “the great bull market in bonds” which means risk adjusted interest rates have dropped continuously over that time and are now at record lows bouncing off the 0% lower bound. (And in some cases managing to go negative…) I.e. the ratio of capital to yield has dropped over time so that lower yield projects have become viable but also the ratio of capital to wages has climbed considerably.
As bargaining power for workers has decreased, but the economy has grown enormously, there are now simply more jobs available at the median wage that are worthwhile for employers. If there is an abundance of capital but wage costs haven’t risen then companies can afford to employ more people and as bargaining power is so low the unemployment rate at which inflation rises above 2% is simply lower. The ratio between capital stock and the median wage has been altered substantially, lowering the natural rate of unemployment.
Over a thirty to forty year period the amount of the population that can be employed at the median wage without spiraling inflation has increased. This shouldn’t be a huge surprise- it was the explicit policy goal of post 1970s politics and monetarist theory. The way unemployment has been lowered without inflation rising is by holding down wages more or less permanently. This has, of course, been done through union busting and the advent of a more part time/gig economy norm along with slowly eroding workers rights but it is just as important to recognise that as monetary policy has been used to hold down inflation it has done this by ensuring there is always a sufficient threat of unemployment to stop workers having unchecked negotiating power on wages.
This result is interesting because this combined with the 2008 financial crisis and the 1970s inflation crisis neatly encapsulates these duel and mutually compatible criticisms that Keynesianism and Monetarism make of each other. Full employment reached via demand stimulus, causing private sector jobs, seems to eliminate any restraints of bargaining power for wages while not solving the problem of limited capacity in the economy to meet that demand. Meanwhile lowering the long term rate of natural unemployment via monetarist policies has eliminated the benefits of lower unemployment. One path leads to almost unchecked negotiating power for employees and other leads to almost none.
Modern Monetary Theory and full employment of labour in the context of Chapter 22:
Recently a Keynesian/Minskyist leaning branch of economic theory has emerged called Modern Monetary theory that attempts to reconcile this problem. Confusingly MMT theorists sometimes appear to disagree that the NAIRU/natural level of unemployment exists while simultaneously offering a solution to that problem. This solution is for the government to create a jobs guarantee program and act as an employer of last resort. MMT argues that governments that spend in their own currency are effectively unlimited in their spending power except for inflationary effects. They are not fiscally constrained, in the sense of only being able to spend the money already in their bank account, they are instead constrained by the real world resource limits that limit the current supply capacity of the economy. They therefore argue that the government can always afford to employ whatever labour the private sector chooses not to employ.
Where this differs from either post war Keynesianism or Classical Economics is in how it deals with the excess labour supply in the economy. Classical economics posits that at a sufficiently low price all labour will be purchased by the market and that real wages must simply fall low enough for everyone to be employed. This conflicts, to some extent, with the natural level of unemployment. This is often critiqued because as wages drop demand also drops and thus the demand for labour drops creating a vicious cycle but there is another, less discussed issue.
The natural level of unemployment implies that whatever arbitrary amount of labour is currently available is all worth employing (given a sufficiently low price). This is not a theory we apply to other resources. In a traditional demand/supply model we assume output can be altered so as to meet the equilibrium between demand and supply. This is not the case in a theory where all available labour can be bought as here the equilibrium point must shift to match the preexisting “output” amount of labour (i.e. the current supply of labour). This would be akin to arguing that the price of land or raw materials can fall indefinitely and it still be worth mining those resources or utilising that land. This is not the case however, it may be the case that the price of some land or raw material falls to zero and it is still not worth converting that land to productive use or mining the raw material from the ground. This is a zero lower bound problem and the same is true for labour. It may be that at a given level of demand the price of the last marginal unit of labour would have to be zero or negative before it was worth employing that unit of labour. The same is true of any input in production, there is no guarantee that an arbitrary amount of it will all be useful at the current level of demand. This doesn’t matter for land where deserts can be left fallow and unused or oil to expensive to extract can be left underground or an excess factory no longer of use can be torn down. With labour however there is no such option.
Keynesianism as practiced effectively increased the demand curve for labour until the value of any and all labour available was greater than zero. This, as discussed previously, appeared to cause inflation. A jobs guarantee attempts to solve this by purchasing the unused excess labour rather than pushing the whole demand curve upwards. This is believed not to be inflationary because it doesn’t confer an indefinite level of negotiating power to workers but instead merely sets a floor beneath which they do not fall. This avoids the escalating wages problem while providing a Keynesian stimulus when needed and withdrawing that stimulus automatically during the recovery. (As workers find it more profitable to work for the private sector than for the government at the minimum wage.)
There are some concerns though both morally, is it acceptable to force people to do make-work jobs in order to claim what are, effectively, unemployment benefits? And then also economic concerns, if it is possible for employing an individual to yield less value than it creates are these jobs not an active drain on real world resources that could be put to more productive use? Conversely if jobs can be found that are worthwhile then surely it is worth offering that job at that price regardless of whether there is a jobs guarantee or not?
Where this is interesting in its relation to chapter 22 is where Keynes argued that
“But at present the evidence is, I think, strong that the majority of individuals would prefer increased income to increased leisure; and I see no sufficient reason for compelling those who would prefer more income to enjoy more leisure.”
No doubt proponents of a jobs guarantee would point to this as a quote in their defence but Keynes’ statement is caviated “at present, I think”. While Keynes is undoubtedly a great modern thinker it seems somewhat spurious that his opinion of whether people would prefer more leisure or work is terribly relevant in a modern context. He also says “income” not work. The reality is that these are individual preferences that are unknowable. If they were knowable then state planned economies would be the successful norm, they are not the norm precisely because it has been found to be more efficient to allow individuals to make demands based on their own preferences. If during a downturn a person is unable to get work and must choose between a job approved on the jobs guarantee scheme and no benefits at all then no doubt that person will take whatever job is required. There is no way of guaranteeing though that, however the jobs guarantee scheme may be constructed, this is the most productive use of their time. If, conversely, an unemployed person may choose to claim an equivalent benefit rather than a jobs guarantee job then the system is actually a universal basic income with a government sponsored volunteering scheme.
What the MMT approach does achieve is creating a floor in demand that combined with a monetarist interest rate policy may make investment sufficiently attractive, even during a slump, as the guarantee in demand stability makes borrowing at the new low interest rates worth it for investors and producers. It does not solve the issues Keynes had with fluctuating interest rates described in chapter 22. It does help somewhat by being an automatic fiscal stabilizer that contracts during a boom and expands during a bust which should reduce the magnitude of interest rate changes necessary to achieve a monetarist style management of the economy but it doesn’t solve the problem completely.
A second important point made by MMT theorists is a point raised by Minsky around balance sheet dynamics in the economy. All flows must balance out overall and therefore the government sector deficit is the private sector’s surplus and vice versa. The argument therefore goes that in order for companies and private citizens to make a profit the government must run at a loss. Minsky’s analysis is more about how this relates to fiscal Keynesianism in relation to offsetting booms and slumps but MMT theorists point out that it is important that over the economic cycle as a whole that the government runs a deficit so that the private sector can run a surplus. This means an ever expanding government debt or an ever increasing monetary supply.
If the government were to run a surplus and pay off debts then the financial sector would have more reserves but no place to put them (as the total quantity of available government debt has obviously shrunk). This will decrease interest rates and allow more private sector borrowing. This is exactly the argument Monetarist (often referred to as neo-liberal) economics would make in favour of government paying off debt, but the argument MMT theorists are making is that as the size of government shrinks the private sector is increasingly reliant on loaned funds rather than government demand. These can be loaned funds to consumers or loaned funds to businesses. Central banks also only inject reserves into the economy via the financial system and so indebtedness must increase to take advantage of those funds. Therefore the private sector can only expand if either government deficits are sufficient to support that expansion, the central bank injects cash directly into the economy or the level of private sector debt increases. I.e. in practice for the economy to expand either private sector borrowing or government sector borrowing must go up. This creates unstable debt bubbles like the one that led to the financial crash and great recession. The government doesn’t have to run a surplus to cause this, it merely needs to run a deficit too small to keep pace with economic expansion. Because government spending is only limited by inflation, argue MMT economists, the government can and should always spend up to its inflation target whatever deficit that creates because that is the only way for the economy to expand to the full extent it can (without breaching that inflation target) without large private sector debt bubbles. This should, according to MMT theorists, help alleviate both inflation worries and Minskeyesk worries around unstable private sector debt instruments.
A few notes on Neo-fisherism:
Irving Fisher (for whom neo-fisherism is named) is credited with pointing out that the zero risk interest rate (for example short term government debt) should follow the formula:
Nominal interest rate = expected inflation + competition for capital
This isn’t hugely disputed but the assumption that neo-fisherites add in most definitely is disputed, and that is that under modern financial systems the competition for capital is frictionless and is therefore zero (or at least constant). I.e:
Nominal interest rate = expected inflation
Therefore if the central bank raises interest rates it will also raise inflation and if it lowers interest rates it will lower inflation. The idea that competition for capital is always zero or constant is questionable but it may be that some “neo-fisherite” forces exist that are capable of rendering traditional monetarist policy less effective. (Rather than meaning traditional monetary theory doesn’t work at all.) Two possible mechanisms that might work to create a neo-fisherite effect are discussed here. The argument here is not that these forces are everything that’s going on in an economy but instead that they may well be real forces that do act in the economy and have real effects that can make interest rate based monetary policy less efficient.
First of the effects is about interest rates themselves and distribution of the money supply. Traditionally monetarism works on the view that:
Price level = (money supply × velocity of money) ÷ nominal GDP
Monetarists assume velocity of money is constant (this is also something that may be a rule of thumb that is generally accurate but really describes a tendency rather than an iron law) and therefore money supply can simply be altered as GDP changes so that prices rise at a low and steady level.
Price level overall covers the entire economy but crucially different types of prices can rise at different rates. It is perfectly possible for prices of financial assets to rise faster than prices of goods and services in the shops. (Something we’ve seen a lot of during the period of QE.) This does make some sense in the context of interest rates. As prices of financial assets go up yields and interest rates correspondingly go down. I.e. their price level has risen. If the transactions happen at the same pace, i.e. velocity stays the same, then more money will be required for each purchase and with the same volume of purchases that will therefore require greater utilisation of the money supply. The money supply could, of course, be used in the purchase of both financial assets and physical goods but in order for the price of physical goods to rise because of that dual use the velocity of money must also increase. It is therefore perfectly possible for money to get trapped in the financial system whereby it enters the financial system and is used to buy a financial asset which drives up the price of other financial assets which means all subsequent purchases of financial assets require more money for the same transaction, i.e. a rising price level for financial assets. Anytime interest rates go down the prices of financial assets must go up and therefore there is less currency available for day to day transactions. This would logically limit the inflationary effects of QE. That fits quite well with what we’ve seen happen.
It’s important to note here that there are two types of financial assets- one where this applies and one where it does not. New financial assets that are used to finance real investment will also push on demand for goods and services to fulfil the investment goals. Purchasing existing capital assets from other investors, however, does not increase demand for anything except financial assets themselves. Therefore the bigger the proportion of existing capital stock to new investment the greater this effect will be.
Mechanism two is really just a function of modern financial systems. If interest rates go up then banks do have to charge companies more for loans, however, that applies to ALL companies that are taking out loans and happens largely all at once. Companies simply have to accept the new terms or immediately go bankrupt and so many will accept higher fees in the hope that they’ll be able to put up prices without losing customers. (They effectively have no choice.) They then discover that they haven’t lost any customers because their competitors have also been forced to raise prices for the same reason. The process works in reverse with lower interest rates. As loans become cheaper companies realise they can lower prices in an attempt to undercut the competition and attract customers, this drives down prices overall. This effect can be considered as analogous to oil prices (which affect almost everyone) rising and pushing up prices or falling and allowing price competition.
Between these two mechanisms it’s certainly possible to imagine neo-fisherian forces but it would be misguided to simply assume neo-fisherites have discovered a universal truth.
Firstly most economies buy things from places with different currencies. Even national currencies aren’t total monopolies and so exchange rates do matter. These effects may push on prices but if trade is pushing in the opposite direction then their power simply won’t be absolute. This might partly explain why QE was so ineffective- all the major economies did it at roughly the same time and so it didn’t have the exchange rate effects you might expect if only one economy had done it.
Secondly the assumptions here are questionable- is velocity of money really always equal? Equally capital friction may be very low with current banking systems but can we assume it’s always zero or at least stable, especially in the short run?
Thirdly just as asset prices rise with lower yields the amount of funds an organisation can borrow while paying the same interest increases. This is the classic monetarist reason for using interest rates!
Finally, lag- contracts don’t all roll over at the same time and neither do loans. Therefore companies will not all be hit perfectly symmetrically by an interest rate hike. Besides which companies obviously aren’t all equally indebted/in need of capital.
Therefore it seems reasonable to argue that in the short run raising interest rates could temporarily jam up the system for some companies and not others and therefore slow the economy. Equally a sudden drop in interest rates might give a temporary boost. They are a short term tool that can be deployed quickly, unlike fiscal policy changes. What this does show though is that interest rates may well work in the short run but it’s far from clear that they are a clean and elegant method of macroeconomic stabilization and probably aren’t perfectly reliable. This is the point Keynes was making in chapter 22. Even if modern banking systems ensure capital availability in the long run Monetarism relies upon the system failing to achieve this in the short run and thus, in effect, central bank policy alters scarcity of capital via rent on money in order to ease or restrict financial conditions in the economy. It is macroeconomic stabilization via the manipulation of the rate of economic rent in the short run.
This can be seen once we remove the assumption that competition for capital is always zero (or constant) and try to achieve Keynes’ aim of low and stable interest rates.
If Nominal risk free interest rates are successfully held stable while inflation is rising then competition for capital must decrease to achieve this. If more economic growth is possible then these are ideal conditions for that growth to occur- cheap and available capital with rising prices encourages companies to borrow to take advantage of the possible economic growth. That growth offsets inflation and slows down the price rises. If further growth can not be achieved however, and the money supply grows, then inflation must rise. This is the scenario monetarism is trying to avoid. In order to do this however it must more than just raise interest rates to match the new higher inflation (as this would just create the rising inflation and rates concept from neo-fisherism). It must instead raise interest rates beyond the rise in inflation so that scarcity of capital is also increased. That is what will ultimately, hopefully, choke off inflation.
Conversely lowering interest rates during a slump must also outpace the drop in inflation from lack of demand. This drops competition for capital and encourages a recovery. If rates are instead held constant then the price level of goods is unable to justify investment at that level of interest/economic rent.
Keynes refers in chapter 22 to the rate of interest rising after a panic but here he is referring to the general level of interest and not the level of interest specifically on safe and liquid assets. Nominal interest on an asset that carries some risk should be inflation expectations + competition for capital + risk. The reason why Keynes is describing interest rates rising is because the perception of risk has increased, but in the context of Fisherite risk free assets the reverse is true. As capital is transferred from riskier assets to risk free and liquid ones competition for capital goes down on those risk free assets. Equally inflation is likely to fall as consumers become more hesitant with spending unless there is some shock to the productive capacity of the economy or government policy is being used to stimulate demand. For these reasons interest rates on safe and liquid assets should fall also. To keep interest rates constant would require pushing up either inflation (presumably via demand) or competition for capital.
Fiscal monetarism- a deliberate contradiction in terms:
Economic growth under the post war consensus and Monetarism (neoliberalism) is actually broadly the same. This really shouldn’t be the case because we’ve been nowhere near the full employment of labour for large sections of the monetarist policy period without a substantial drop in GDP growth per capita and so what this implies economically is that it’s completely possible to achieve maximum potential output of an economy without achieving full employment of labour. This is an entirely separate issue from the social consequences of increased unemployment- that’s an important but separate question, this is just about the economics. So when Keynes talks about it being too early to accept the point made by DH Robertson that:
“...full unemployment is an impractical ideal and that the best that we can hope for is a level of employment much more stable than at present and averaging, perhaps, a little higher.”
It may be that Robertson is closer to the truth than Keynes might like to admit. Perhaps either with higher inflation or lower wages or both something approaching full employment may be possible but we do not need it to maintain economic growth and achieving it may either destroy the economy via inflation or else so impoverish the average person that it can’t be seen as a positive goal.
So given that knowledge it is possible to argue that there are reasonable grounds for updating Keynes. What that doesn’t mean is that monetarists are right about everything. Clearly there is considerable frustration and disquiet with the ‘neoliberal’ economic system. Recovery from the great recession has been painfully slow both for GDP and for wages. Inequality of wealth and income appears to be substantially reducing equality of opportunity and wiping out economic gains for large swathes of people while enriching only a few.
The aim, therefore, is to develop a system that takes into account the points made by both Keynes and Friedman, that keeps interest rates and inflation low while not having an economy that overheats or drops wildly below maximum potential output with policy tools for both the short and long term. Modern Monetary Theory and Minsky provide a helpful starting point in how to avoid ever eroding wages without triggering spiralling inflation. It can be further improved, however, by reintegrating some of the insights from chapter 22. Particularly around redistribution and interest rates.
Under the MMT rubric of national finance the government (in the example here acting through the central bank) can be seen as analogous a power plant for the economy. It creates money (at least in terms of reserves), that money then circulates and can be destroyed later, if necessary. That’s like the electrons in a circuit that are pushed out into the circuit, circulate and are then brought back. It’s not a perfect analogy but no analogy ever is. In this description traditional monetarism is a little bit like adding a variable resistor to the circuit- if too much power is flowing through the circuit we can increase the resistance (interest rate) to calm things down and stop the components in the circuit overheating. (Or the economy from overheating.) It’s inefficient but it works.
There are four additional tools that we might add to a central bank’s toolkit, based on chapter 22 and some insights from Henry George, to achieve the same end more efficiently. The first two deal with achieving the low but steady interest rate. The second two with integrating Georgism/single tax theory into central banking to achieve aims Keynes lays out in chapter 22.
-The central bank could be allowed to issue a bond with no end date.
Instead it is effectively just a promise to pay the coupon for an indefinite length of time. (It’s therefore really an annuity but it is easier to conceptualise as a bond.) The coupon (interest) on the bond is important because the interest would be a perfect fisherian product, there is no risk of default. Only the risk is inflation plus the competition for capital.
So if the bond has a 3% coupon and the central bank manages to ensure that the market value of the bond remains at 100% of its initial value then inflation plus competition for capital must add up to 3%. Inflation, therefore, must be low as long as competition for capital isn’t negative. The new target therefore is a 3% yield on these bonds rather than an inflation target of 2%. 3% is somewhat arbitrary, but then so is the 2% inflation target. 3% was chosen because it’s low but gives a little more flexibility above the lower bound than 2% does.
-Expanding the use of reserve/capital requirements.
By altering how much cash and how many of the above bonds commercial banks are required to hold it would be possible for the central bank to force up and down the market price of the bonds so that they hit their 3% yield target.
In an economic slump when investors are desperate for safe assets, the market price of the bonds will be above 100% of their nominal price, i.e. the yield will be below 3%. By lowering reserve requirements banks within the system would be able to sell the bonds they own to investors and continue to loan out funds. If appetite is strong enough they can even purchase new bonds from the central bank. This increases the supply available in the market and decreases the demand forced on banks by the reserve requirements. This pulls the price down and pushes yields back up towards 3%. As reserve requirements drop then the amount of funds the banks are able to loan out from the same reserves increases and it should therefore increase the supply of loaned out funds. Assuming there is sufficient aggregate demand to ensure the investments are profitable.
Conversely during a boom the market value of bonds will fall below their nominal value and effective yields will rise as private opportunities offer higher yielding assets. Reserve requirements can then be raised so that investors who are feeling buoyant and want to sell their bonds can do so due to the higher requirements placed upon the banks. This drives prices back up until yields are back around 3%. It also causes investors to have cash to make investments during the boom but restricts the banks from doing so, effectively transferring risk from the taxpayer backed banks to individual investors willing to take those risks. This also means more of the money supply is required for the same financial market transactions.
Using an annuity/bond for this is helpful because it can be held by both banks that have accounts with the central bank and other entities, unlike central bank reserves which can not. The effect is somewhat similar to increasing reserve requirements while keeping interest on reserves at 3%. This is how the system is able to mimic traditional monetary policy without raising rates- rather than encouraging banks to save via higher rates the central bank simply forces them to with higher reserve requirements. This is the same effect as if private individuals had switched from consumption to purchasing existing capital assets. While taxes can fight inflation by pulling money out of the economy and thus reducing the money supply (if the government chooses to simply not spend what it taxes in) they can not switch spending from consumption to the purchase of existing capital assets. Reserve requirements for these bonds/annuities can and because doing so raises the price of existing capital assets it does so without increasing interest rates.
To go back to the electricity example these two tools together are like adding a battery to the system that stores up power when too much is being generated and discharges that power when there's not enough. It is not just the central bank bonds themselves that are the battery but the entire stock of existing capital assets that is the battery because the price is raised or lowered for the entire capital stock. As yield doesn’t drop below 3% though investors can not ultimately see the price of their assets rise except by higher incomes and to have real terms price rises they must have a real terms increase in incomes on the assets. This is better than interest rates because rather than increasing resistance to 'put sand in the gears’ instead the system is storing excess capacity to be used at a later date. This has echos of a Keynesian government borrowing during the hard times and paying it back in the good times where the inner banking system attached to the central bank sells bonds out during hard economic times and buys them in during the good ones. The evidence from the post war era suggests that economies will usually achieve long term maximum potential output if inflation and interest rates are consistently low and so this system would likely do at least as good a job as the current monetarist system.
This should achieve Keynes’ desire in chapter 22 for a low and stable interest rate that doesn’t need to be raised to stave off a boom and inflation. The way the reserve requirements are structured demand deposits are still cyclically controlled it’s just this isn’t done by interest rates and it ensures that during a boom credit money created by banks will find it difficult to switch from use in purchasing financial assets to buying real world goods as doing so would affect yields, triggering a central bank response.
It is possible to argue that the reason quantitative easing didn’t work, with inflation staying flat and the recovery being sluggish, is because the prices of financial assets can rise either by a growing economy OR a lower interest rate. A commonly used phrase in the post crash financial world has been “the search for yield” i.e. as money was pumped into the financial sector average yield on financial assets simply fell. This didn’t actually result in a recovery and the money injected into the system simply circulated buying ever more expensive financial assets.
It is probably prudent to consider whether the first two tools alone would solve the problem that current monetarism experiences. What would happen under the equivalent scenario?
Yields can’t drop below 3% because the central bank will always issue risk free assets yielding 3%. That doesn’t solve the problem though, it just displays it more clearly. Instead of the cash circulating to buy ever more expensive financial assets the cash simply gets deposited back with the central bank in exchange for the 3% annuities/bonds. So we need a way for the central bank to get cash directly to consumers in the economy. The obvious solution, therefore, is a ‘helicopter money’ drop via a citizen’s dividend/basic income. The last two tools are designed to create a clear and transparent system that achieves this without causing excess inflation. This is the other key aspect of chapter 22- redistribution. This is where the integration of Georgist ideas comes in.
It is important to know what tool will be available to remove any excess reserves should the central bank overshoot on redistribution. The reserve requirements will help but this is a little bit akin to financing all government spending via debt. What is needed is a tax. The obvious choice is land value tax. Land values already fluctuate with the economic cycle so the tax will be relatively easy to sync up with macroeconomic policy and LVT targets rentier income and so shouldn’t create deadweight losses or be passed onto consumers. Further it is very difficult to avoid and can be raised to whatever level is required to achieve the objectives.
-LVT that the central bank would set with a target of 1% to 2% land value price inflation year on year.
Why 1% - 2%? It means land values should rise at close to but below inflation, meaning that any real economic value would have to come from productive use of the land but, vitally, land values would tend upwards reducing the likelihood of negative equity and thus avoid defaults on mortgages. During economic booms land values would rise faster and so the tax would rise twice- both because the percentage of land value called for in tax would rise, as the central bank seeks to curb land value price rises, and also because the land value itself is rising and thus any given percentage of that, now higher, value is a larger nominal amount. During recessions the reverse is true, land values will tend to fall and so LVT rates would have to be lowered to halt the loses.
This is austerity during the boom and stimulus during the bust. A standard Keynesian fiscal policy. This is, however, a slower policy tool than interest rates but run alongside the bond reserve requirements acting in the short term the LVT could be a very effective medium and long term tool.
This sets the central bank up so that it can have helicopter money with more than enough tools to fight inflation if and when necessary.
-A basic income level payment paid to every citizen.
This provides a consistent safety net both for the individual and for demand stability during recessions even with sticky wages. If inflation is in the 2% to 3% band, preferably at around 3% to minimise economic rents, then the UBI should be allowed to rise at 4% a year so that living standards consistently rise. If inflation is running too hot even with the bonds/annuities and LVT then the central bank should be allowed to lower the rate of increase for the UBI with the aim of keeping the rate of increase above inflation, if possible, and if not to never have a rate of increase below 2%. This gives flexibility but errs on the side of every citizen getting an increasing share of economic growth. If inflation is stubbornly low (i.e below the 2% mark) and this is consistently true even as LVT and bonds kick in then there’s really no economic reason not to increase the UBI to any level that pushes inflation back up into the 2-3% range. The only risk of too high a UBI is inflation. If no inflation is forthcoming then greater raises simply mean greater prosperity for all. The way this is “financed” is really the tools that ensure excess cash is removed to fight inflation. With LVT being a core part of this it is quite legitimate to argue that the basic income is coming not from “other people’s hard work” but out of unearned rentier income. This is simply taking the tools laid out in Georgist thinking that were intended to create a fiscal system that taxed unearned economic rents and distribute them more fairly and applying those tools to central banking. Democratic government may choose to layer taxes and services on top or it may not, from an economic point of view the system should work either way.
Here it is useful to compare and contrast with a universal jobs guarantee and cover how a UBI can achieve the same economic result in a more resource efficient manner.
A job guarantee does have one major advantage in that it is an automatic stabilizer- i.e. it kicks in during downturns and is slowly withdrawn as the economy recovers. (I.e. the private sector can afford to employ more people and so hires them away from the jobs guarantee.) This is true but doesn’t solve the problem of misdirected resources used in make work schemes created to ensure the jobs under the jobs guarantee.
Most jobs guarantee advocates argue that no make-work jobs are necessary, but this seems hard to square with the practical experience that we can be below full employment and yet at maximum potential output. There seems to be no innate reason why every additional piece of work done would be productive and worth the real world resources required to facilitate that activity. It’s just assumed that additional jobs mean additional useful output this isn't necessarily true for the reasons discussed earlier. Again it's not clear why individuals themselves would not be best placed to decide how to use their time and resources and therefore the economic system merely needs to ensure they have the necessary resources available to do so.
UBI is also more nuanced than it first seems, as it is net government transfers that matter and not whether each tool is, in and of itself, an automatic stabiliser. If people lose their jobs in a recession and rely on the UBI instead then although the UBI amount they receive is nominally the same, the net amount they receive is not because their tax bill will change. This is actually an argument very much in line with MMT which argues governments can alter taxes and spending independently to achieve objectives (such as full employment) as long as they’re aware of the trade offs of doing so, such as possible inflation.
This is more easily understood by comparing UBI (universal basic income) with NIT (negative income tax). They are often argued to be very different, UBI paid to everyone and NIT only paid to those on low incomes. However they have been demonstrated, many times, to be the same and this helps show how UBI can be used as a component of an automatic stabilizer even though it is not one in and of itself.
To reiterate, MMT argues that monetarism solves inflation by essentially creating unemployment. When the economy becomes overheated and workers demand higher wages but are chasing the same maximum possible number of goods. This causes spiraling inflation and workers having to demand even more pay, but when the monetary authority raises interest rates workers lose bargaining power as companies face tighter financial conditions and must hold down wages or lay people off. MMT refers to this as a “buffer stock of unemployment” MMT argues for doing substantially the same thing except when workers are laid off there is a jobs guarantee scheme where the pay rises at a preset, steady rate (presumably close to but slightly above the inflation target) so workers still can’t demand ever rising wages causing ever higher inflation but they also aren’t left unemployed. They refer to this as the “buffer employment ratio”- BER
UBI is paid to everyone, not just the unemployed, but nevertheless it is also a preset amount that steadily rises (presumably at or just above inflation) that individuals may end up reliant on if they lose their job in a downturn. It is necessary to understand that taxes render UBI and NIT functionally equivalent. (A negative income tax works by a different mechanism, below a set income you get progressively more money from the NIT scheme and above that set value you pay tax.) This is best shown by example.
For simplicity let’s posit a UBI of £10,000. For this first example there’s a flat tax of 50% but it also works with progressive taxation.
UBI of £10k with a flat tax of 50%:
- £0 income from work plus the £10k UBI = a total of £10,000 after tax income.
- £10k of income from work taxed at 50% plus £10k UBI = £5k income after tax + £10k UBI = £15,000 final income.
- £20k of income from work taxed at 50% plus £10k UBI = £10k income after tax + £10k UBI = £20,000 final income.
- £100k of income from work taxed at 50% plus £10k UBI = £50k income after tax + £10k UBI = £60,000 final income.
Now the same thing with NIT of 50% below £20k and 50% tax above £20k:
- £0 of income from work is £20k below the threshold of £20k and 50% of £20k is £10k so the final after tax income is: £10,000.
- £10k of income from work is £10k below the threshold, 50% of £10k is £5k and £10k income plus £5k from the NIT comes out to £15,000.
- £20k of income is at the threshold so no change. That’s £20,000.
- £100k of income is £80k above the £20k threshold and that £80k is taxed at 50% so the after tax income = £100k - £40k = £60,000.
It is also possible to phrase the exact same scheme as a UBI withdrawn at 20p for every pound you earn, alongside a tax of 30% up until £50k and a 50% tax thereafter. They’re all the same thing mathematically speaking and it’s possible to construct any arbitrary tax scheme and show it as a UBI + tax or as a NIT and they will yield the same results for the same income.
What all this is showing is that NIT is really just calculating the net effect of UBI plus a particular tax regime. That net effect is the final result of either a UBI OR an NIT and while UBI is technically paid to all and so does not alter with the economic cycle the net amount does change and does fluctuate with the economic cycle. That means that the net result of a UBI is an automatic stabilizer and acts in the same way as a jobs guarantee in that respect, it just skips over the work requirement. To put it another way- net government outlays are what matter and not whether each individual outlay is a fiscal stabilizer. Shifting people onto a UBI is an acceptable substitute for shifting them onto unemployment or a jobs guarantee scheme.
MMT reminds us that the costs of government spending are really about controlling inflation and crowding out effects. Inflation can be dealt with with taxes and the tools discussed above, and crowding out doesn’t apply to a UBI when interest rates are held constant. The reason being that the government isn’t buying up and using labour or resources from a UBI, it’s simply shifting who is spending the money, but it’s still very much money spent in the private sector by private citizens. The consistent interest rates via bonds reserve requirements (de facto forced savings) ensure that redistribution effects don't alter the balance between consumption and investment. The crowding out effects only come if government borrowing pushes up interest rates and crowds out investment capital. (This is true of any government transfer payment, it is simply cash spent as the recipient desires. Crowding out does not apply.)
That means taxes can be dealt with entirely separately to the basic income rather than amalgamated into one as in NIT. LVT is particularly useful here as it follows the economic cycle and stops the gains of a UBI simply being sucked up by landlords.
Having a UBI in no way stops people from doing the work, volunteering or education they would do under a jobs guarantee if they wish. Instead it gives them the choice to do it or not. It allows them to make the decision about what is the best use of their time rather than an official who decides what counts as a job under a jobs guarantee and what does not. It also in no way stops there being programs to help people find job, voluntary or education opportunities nor does it stop benefits programs for those who will require additional help, like the disabled.
A jobs guarantee means finding people work but if this work is worth doing at the wages set under a jobs guarantee then there is no need for a jobs guaranteed to create that job- it's independently worth paying for. Anything that isn't worth paying for outside of a jobs guarantee is simply make-work and forcing people to do make-work when they may be able to find other ways to fill their time that they consider more worthwhile seems to be a waste of time, energy and resources.
Finally there must be rules for the interbank lending rate and bank collapse. Through open market operations buying its own bonds/annuities (or government bonds if necessary) the central bank can always lower interbank overnight borrowing costs below, say, 4% (Ideally it would hover around 3%). If the interbank rate falls below 3% then banks will be increasingly likely to swop reserves for central bank bonds/annuities pushing interbank lending rates back up. This ensures that the central bank can keep costs low generally while allowing any individual bank to fail if it is poorly run. Direct lending to a bank should only happen if it goes bankrupt. In that instance all its demand deposits should be considered valid and transferable but all equity should be wiped out and new shares issued, but with dividends banned. Any liabilities that the bank owes (other than demand deposits) should be swapped for a 3% annuity from that bank just like the ones from the central bank. These should include a repurchase agreement so that the bank can buy them back at full price if and when it can afford to. Only then should it be given access to direct central bank lending at 4%. Once the bank can borrow on the interbank rate at less than 4% consistently it can exit the bankruptcy mechanism and pay dividends again but it will no longer be able to access central bank loans at 4%. This is designed to make deposits very safe without creating moral hazard.
None of these four tools enforce a particular style or size of government and should be flexible enough to cope with a very wide range of situations both economically and constitutionally. By taking the Henry George concept of the one tax state and applying it instead to the central bank room is left for democratic governance to choose anything from small government libertarian capitalism to Social Democracy, Radical Liberalism or Democratic Socialism. This could conceivably help solve problems in the Euro area or Japan while working just as well in the US, the UK or in less wealthy nations like India or Brazil. It is intended to be able to combat something like the ‘08 financial crisis or 1929 crash, without having to bail out the banks, but it should be equally applicable to high inflation scenarios like the 1970s. It achieves the aims Keynes sets out in General Theory as a whole, and chapter 22 in particular, without running afoul of the inflationary spiral problems of the 1970s. By accepting the mutually compatible criticisms Keynes makes of “purely monetary” policy and Monetarists make of post war Phillips curve keynesianism hopefully this paper provides a solution that takes the best of both without the flaws of either.