Recent Posts

Wednesday, October 11, 2017

MMT And Automatic Stabilizers

The recent internet debates about Modern Monetary Theory (MMT) have been interesting, but the various critics of MMT have largely missed the elephant in the room: automatic fiscal stabilisers. In my view (which may not reflect the official "MMT Party Line"), one of the keys strengths of MMT is that it is largely built around the importance of automatic stabilisers, and institutional details. The conventional view is to acknowledge the existence of automatic stabilisers, but otherwise pretend that they have no effect on the economy.

This article was triggered by the following comment by Neil Wilson on a previous article on my site:
What is interesting about Wren-Lewis, as well as your [another commenter -BR] comment, is this seeming inability to see the automatic stabilisers (particularly the spend side auto-stabilisers) - which require *no* human beings involved at all. They just work - instantly, spatially, automatically.

Money: Still Not Abolished

As one might expect, some of the internet discussion veered off into discussions of money. As I would expect, these discussions had little value-added. This is yet another example of why money needs to be abolished from economic theory. (To be clear, abolishing money from economic theory is my personal crusade, and is not a feature of MMT. However, MMT is not greatly harmed by its abolition, which is why I am in the MMT camp. Also, many people do not understand what I mean by abolishing money from economic theory, to which I respond: my book is available at all major internet booksellers.)

We need to drop our mystical debates about the origins of money (which appears to occur before the advent of writing as we know it) or semantic arguments about the definition of "money," and just ask ourselves: why do "modern" states use money?

The answer is straightforward: to provision itself. The alternative to using money is to directly requisition goods and services -- such as press-ganging unfortunate bystanders into the Royal Navy. Direct requisition was essentially used during the World War II command economy (in many combatant nations), but we prefer to avoid its use during "peacetime."

In most developed economies, governments issue their own money, and they use this same money to pay for goods and services. In some countries, an external currency is used, but this generally only makes sense for small countries, or badly-managed ones.

The reason why governments need to impose taxes is that they need to reduce demand in the private sector to make goods and services available to the government. Alternatively, they need to use taxes to create a demand for the currency they issue so that it is valued by the private sector (and hence, the government can buy real goods and services for its fiat currency).

We can now circle back to automatic stabilisers. One of the insights from MMT -- that is completely missed by most critiques -- is that the "value of money" that matters to the government is the exchange ratio of real goods and services the private sector is willing to provide in exchange for money. As implied in my article on price levels -- there is no guarantee that this measure of the value of money matches any other measured price level. The fact that this is divorced from how households value money is probably the cause of a great amount of confusion. From the standpoint of economic policy, the government's view matters, not households'.

One of the failures of modern policy is that governments ignore their role in the setting of the price level. There is an unquestioned belief in the supremacy of market prices. Since government spending policies adapt to prevailing private sector prices, it easily becomes an engine of inflation. (Rising private sector prices are met by increased government spending, perpetuating excess demand.) In other words, the existing institutional regime in government finance is an automatic destabiliser with regards to the price level. This entire possibility is swept under the rug by conventional approaches to macro analysis. Since the possibility is assumed out of existence, the MMT insight is ignored.

Automatic Government Finance Policy

I do not have a reference handy, but I believe the following three rules would summarise Warren Mosler's suggested reforms of government finance.
  1. The rate of interest on settlement balances (reserves) at the central bank are 0%.
  2. Treasury bills (3-months) would be issued at a fixed (annual) yield of 0.25% in whatever quantity bidders want to pay.
  3. The central bank would be institutionally consolidated with the Treasury, and there will be no chance of default on any government cheque. (I am unsure how he phrases this point.)
Under this regime, the percentage of government liabilities held as government debt levels (Treasury bills) would be entirely determined by the private sector portfolio allocation desires. (Which matches up with the description in mainstream or stock-flow consistent models.) Government debt managers would have nothing to do.

Since government debt managers would just be accountants (making sure no funds go astray), they do not need to be staffed by New Keynesian doctorates that are churned out by the Ivy League/Oxbridge.

The Job Guarantee

Arguing that MMT is just a return to the old Keynesian belief in fiscal policy stabilisation is disingenuous. The "brain trust" behind Old Keynesian macro believed that central planners could manage the business cycle through "stop-and-go" fiscal policies. That policy stance was diagnosed by heterodox economists (such as Hyman Minsky) in the 1960s, and is not the core of the MMT approach. Instead, the reliance is upon the automatic stabilisation provided by the Job Guarantee.

One can imagine a hypothetical world where the Job Guarantee essentially eliminates the need to worry about the business cycle from a policy perspective. (Equity investors would still need to worry about the cycle, as discussed later.) Let us assume that all workers are paid a uniform wage, whether they are working for the private sector, or enrolled in the Job Guarantee. (Employers would need to compete on the basis of the attractiveness of the work environment, or non-wage benefits.) In this situation, if a worker lost a private sector job, it would be replaced by a Job Guarantee job with equivalent pay. There would be no need to worry about the household's capacity to meet debt obligations, etc. The total wage bill will always equal the number of people who want a job times the uniform wage.

This set up does not imply that the business cycle would be abolished: wages might be largely invariant, but profits would not be. Firms might rise and fall in line with patterns of fixed investment. That said, why should policymakers care about that? Capitalists earn profits by risk taking; the possibility of failure the counter-weight to their dividend cheques. We would no longer liquidate labour during the downturn, just capital. (To do: insert patriotic appeal to creative destruction here.)

(Of course, we do not live in a world of perfect equality of incomes. The main limitation of the Job Guarantee to act as an automatic stabiliser is wage inequality: highly paid workers will take a severe income hit if they end up in the Job Guarantee programme.)

In any event, there is once again no need for New Keynesian economists to fine tune policy. As a result, it is not entirely a surprise that such automatic stabiliser effects are white-washed out of New Keynesian models. Once again, the insights of MMT are ignored -- because it is assumed they do not exist.

Concluding Remarks

Modern Monetary Theory involves analysis of economic institutions. It seems clear that such analysis is not popular, and so the focus shifts to less important aspects of the theory.

(c) Brian Romanchuk 2017

33 comments:

  1. I think we could all agree that a store owner could pay employees by issuing certificates granting fixed-amount access to store merchandise. Some people might describe this practice as paying employees with 'gift certificates'.

    We can follow this example and compare money to gift certificates. Currency issuing governments can pay for services by issuing money (or 'gift certificates').

    There are no "automatic stabilizers" possible in this economic model. Everything that is done is the result of human decision. There is only good management or failure-of-good management.

    In my view, we can conclude that MMT can be easily (and logically) redefined into a theory that precludes "automatic stabilizers".

    ReplyDelete
    Replies
    1. What an individual firm or household can do tells us little about the government's position. Even a minimal central government spends several percent of GDP in a year. Policy analysis has to be from the perspective of the central government, not an individual.

      Delete
    2. Yes, indeed!

      The utility of the analogy comes from the ability to scale from a business owner's perspective to a national owner's perspective.

      China may be an easy comparison. One party rule allows the Communist party to act like it owns the nation. One individual at the top of the Chinese government (or party) has the ability to make a decision that affects the entire nation. The effect is identical to the more limited role of a private business owner. Only the scale is different.

      Delete
    3. Roger,

      You missed the bit where the store is run largely by robots and can produce enough for all without engaging everybody.

      The Job Guarantee is there to allow everybody to sell their 8 hours a day. That's the auto-stabilisers.

      In a capitalist system the number of slots is variable. You need a buffer stock system to ensure all the hours are sold.

      Delete
    4. Roger, you originally wrote: "In my view, we can conclude that MMT can be easily (and logically) redefined into a theory that precludes "automatic stabilizers"."

      Why would you possibly want to do that? The whole point of my article is that a great many policies do act as automatic stabilisers, and we want to increase their potency.

      Delete
    5. The lesson (drawn from the analogy between store owner and government decision maker) is that issuance of money is controlled by human decision makers. It follows that the size of a Job Guarantee allowance would need to be based on some criteria and approved by human action.

      You have pointed out the difficulty inherent in using a price index as a criteria reference. When we consider that money-in-hand is like having a gift certificate that can be spent anyplace, how does any government decision maker decide how to 'fairly' allocate the monetary resource?

      It seems to me that automatic stabilizers cannot be automatic at all. They become an ownership judgement made by an individual or group of individuals.

      Delete
    6. Roger Sparks,

      I believe that “automatic” means that, once the law is enacted, you do not need to change it or enact new laws. Spending will eventually rise “automatically” during depressions, and decline “automatically” in booms.

      “Discretionary” spending, on the other hand, will not rise or fall “automatically”.

      For example, if a big federal Job Guarantee (JG) was in place (with a big federal budget) that guarantees jobs for USD 7.50 per hour for anyone wanting to work, then it would be “automatic”. During depressions, a lot of people would volunteer to the JG program, and in booms people would quit the JG to work in the private sector. Everything would happen without new laws or government intervention. All you would need was the program in place. Of course that the 7.50 wage is a policy decision.

      On the other hand, during depressions, the government could just start to build a lot of bridges and roads (or railroads) around the country in crisis (or any other thing it deems useful). You would need a law and a budget for every project, and once they were finished, you would need to find some other projects. The government would need to evaluate somehow if the country was going through depressions, maybe monitoring the unemployment rate. As you can see, nothing would be “automatic” in this scenario. Everything would be discretionary.

      Delete
    7. Yes, “automatic” in this case means that there is no need for a technocrat to estimate the current state of the economy, and set policy to lean against deviations. All spending programmes have to be administered by bureaucrats, or funds will go astray.

      Delete
    8. Romanchuck. - Honestly you are just wishful thinkig. Nothing happens 'automatically' when policy is formed an enacted by government. There is a process which involves decisions.
      How would bureaucrats administer any payment of there is not some decision on what the sate of the economy is? And how would any financial administration be carried out if there is no policy.
      Your theory completely ignores the fact that bureaucrats need a certain allocation of money and that money is determined by government, i.e. politicians when they do their budget. That budget is done according to the economic conditions and those conditions are assed by whom? -- Yes, Advice from technocrats of one sort or another whether that be good advice or bad advice.

      Delete
    9. Bureaucrats do not need budget approval for making a payment for someone on welfare or unemployment insurance; they just validate that the person is eligible. In Canada, the unemployment benefits are completely outside the budget process; not sure how other countries do it. At worst, they just increase the budget allocation if spending is about a limit.

      I didn’t invent the phrase, economists have been using it for decades.

      Delete
  2. It’s news to me that conventional economics prior to MMT acknowledged “the existence of automatic stabilisers, but otherwise pretend that they have no effect on the economy.”

    The first economics text book I picked off my book-shelf published prior to MMT is quite clear that automatic stabilisers have a very definite effect. Indeed a “stabiliser which has no effect” is a contradiction in terms surely, whether one is referring to engineering (e.g. stabilisers on ships) or economic stabilisers? (The text book is “An introduction to Macroeconomics” by J.Harvey and M.Johnson, published in 1971.)

    ReplyDelete
    Replies
    1. When I wrote "the conventional view," I am referring to the conventional view in 2017. If I want to include every single possible qualifier in every single sentence I write, this article would have been 10,000 words long.

      To be clear, I did not want to imply that MMT "invented" the concept of automatic stabilisers, just that the mainstream threw the concept down the memory hole (at some point after 1971).

      Delete
    2. Well I just looked up "automatic stabilizer" in three online dictionaries of economics, commerce, etc. Didn't see anything about them not working.

      Delete
    3. Try finding an estimate of their importance in the DSGE modelling literature, which is a lot more relevant to what I am writing about.

      Delete
  3. "The main limitation of the Job Guarantee to act as an automatic stabiliser is wage inequality: highly paid workers will take a severe income hit if they end up in the Job Guarantee programme."

    That is supposed to happen as part of the creative destruction process. 'Highly paid workers' that cannot get another job in the market at the 'highly paid level' are in fact overpaid by definition and part of the capital misallocation. So you get a contraction in the worker side as well as the capital side and whatever cascade effects that initiates.

    But the existence of the Job Guarantee stabiliser allows for a more flexible job arrangement which makes such occurrences less likely to occur.

    Part of the failure of the 1970s was propping up jobs and firms that no longer made any sense under capitalism. The reliance on private firms for job provision and that they would 'do the right thing' was a colossal mistake. If you're going to have a capitalist system, then jobs are protected by the local market for jobs, not by any particular job with any particular firm.

    One of the points of the Job Guarantee is to ensure that everybody is 'talent' in the job market. You get the detrimental effects if capitalism is able to turn workers into 'gig' workers. .

    ReplyDelete
    Replies
    1. My point is that the total wage bill becomes cyclical, which would not happen under the hypothetical perfectly flat wage structure. Since such a flat wage structure is not going to happen any time soon, I am not going to worry about the side effects of that situation.

      Delete
  4. "Since government debt managers would just be accountants (making sure no funds go astray), they do not need to be staffed by New Keynesian doctorates that are churned out by the Ivy League/Oxbridge."

    That may be why such types are not too keen on the MMT analysis...

    ReplyDelete
  5. “I do not have a reference handy, but I believe the following three rules would summarise Warren Mosler's suggested reforms of government finance…..Treasury bills (3-months) would be issued at a fixed (annual) yield of 0.25% in whatever quantity bidders want to pay.”

    The only “reference” I know of which is relevant here is Warren’s Huffington article “Proposals for the Banking System” (link below). In the 2nd last para, he opposes all forms of interest on government debt. He says “I would cease all issuance of Treasury securities. Instead any deficit spending would accumulate as excess reserve balances at the Fed.”

    http://www.huffingtonpost.com/warren-mosler/proposals-for-the-banking_b_432105.html

    Milton Friedman advocated much the same, though Friedman thought interest yielding debt could be justified in war-time.

    I basically agree with Friedman and Mosler: i.e. I fail to see any reason to pay anyone simply to hold onto wads of central bank issued money. Though (like Friedman) I wouldn’t totally rule out artificially raising interest rates in an emergency. Apart from the “war” emergency, another possibility is a serious and excessive outbreak of Greenspan’s “irrational exuberance”: raised interest rates might be a useful tool to help damp that down.

    ReplyDelete
    Replies
    1. He discussed Treasury bills being issued at 0.25% in another article (somewhere). There's very good institutional reasons to follow that model. There are class of investors (e.g., some trusts) that cannot take any risk on their "cash" holdings, and are beyond bank deposit insurance limits.

      If someone thinks we cannot "afford" to pay 0.25%, we can set the level at 0.125%. In any event, I would go in the other direction, and lock the nominal rate at either 3% or 5%.

      Delete
    2. Is there a need for a marketable, transferable instrument to do that though?

      Why not simply increase bank insurance limits. Depositors are not investors in the bank, so why treat them as such?

      There's a big debate as to whether outsourced pension funds add any value to society - particularly when you no longer need the 'wealth effect' to drive things.

      Delete
    3. This Mosler link says Fed should peg Tsy rate:

      http://moslereconomics.com/2013/08/21/macro-update-3/

      "... having Fed peg the entire term structure of risk free rates with, for example, a bid for the entire tsy curve at their target rate ceiling, along with an open ended securities lending facility."

      Regarding job program local governments should be enlisted to determine the types of jobs that qualify for a federally subsidized Job Guarantee and to staff qualified supervisors. So as Roger describes there must be a management function that is not "automatic". Paying the wage bill for the job program, however, would be an automatic fiscal operation carried out by the federal or central government.

      Delete
    4. I would make the T-bills perpetual and redeemable on demand at par. And at 0.24% interest (a round $2 per month per thousand), paid monthly. No partial interest at redemption. Not that they would be popular vehicles, except for the cases cited in which they would be the only feasible one.

      Delete
  6. "Modern Monetary Theory involves analysis of economic institutions. It seems clear that such analysis is not popular, and so the focus shifts to less important aspects of the theory."

    I get that seeing the same money & banking 101 discussion for the nth time is... well, let's be generous and call it tedious. How is that to be avoided though when in most cases the roadblocks or responses to that analysis almost always rest on an incorrect understanding of money & banking?

    ReplyDelete
    Replies
    1. The solution is to delete “money” from economic theory!

      Otherwise, you need to ask yourself: what practical difference does this theoretical dispute mean in terms of real world policy outcomes? The deletion of automatic stabilisers from some descriptions of MMT leads to an obviously incorrect description of the view of fiscal policy.

      Delete
  7. Good discussion! Thanks, Brian and all...

    ReplyDelete
  8. Brian I read your book "Abolish Money (From Economics)" and I agreed with just about everything in it except for the title, because nowhere in it did you make the case that money should not be something that economics should discuss, or try to understand. Now it may be that you have a different understanding of what 'abolish' means, but what about these statements quoted from your article?-

    "However, MMT is not greatly harmed by its [money's] abolition". I would just remind you that MMT is short for 'Modern Monetary Theory' so the word money is in the actual name.

    "We need to drop our mystical debates about the origins of money (which appears to occur before the advent of writing as we know it) or semantic arguments about the definition of "money," and just ask ourselves: why do "modern" states use money?" Why states (and the people therein use money is a good question for economists to ask, but then they would be talking about 'money' wouldn't they?

    And your answer to your question of why they use money I agree with it completely- "to provision itself". And why they use money rather than commandeering goods and services is also a question, but one that has a very obvious answer- Nobody likes their government commandeering their belongings or services. So now we are back to talking about money again.

    "The reason why governments need to impose taxes is that they need to reduce demand in the private sector to make goods and services available to the government. Alternatively, they need to use taxes to create a demand for the currency they issue so that it is valued by the private sector ". Talking about money again and with a very, very different understanding of why currencies have value than most mainstream economists would explain. Kind of important to MMT.

    And I honestly don't know how automatic stabilizers can be reasonably discussed without referring to money either. Government policy that taxes more when income is high take money out of the economy and cool demand, and programs that spend more when income is low put money into the economy and increase demand. That was how they were described when I learned about them years ago, and as far as I can tell that is how MMT describes them now.




    ReplyDelete
    Replies
    1. Money has arisen in most civilisations because of a desire by the ruler / king etc to make tax collection more efficient – at least that seems to be the case from the less than voluminous amount I’ve read on the subject.

      E.g. the Romans collected taxes at least to some extent in the form of agricultural produce when they ruled Britain around 2,000 years ago: not an efficient way of collecting taxes. Thus I suggest that “Nobody likes their government commandeering their belongings or services” doesn’t have much to do with it: nobody likes their government commandeering their money either, do they?

      I.e. the real motive for money is efficient tax collection. Plus of course money increases efficiency in another way: it gets rid of the inefficiencies of barter.

      Delete
    2. Jerry,

      1) The name MMT is a historical accident, and can be renamed. The focus on money by some online MMTers is a tactical mistake.

      2) Automatic stabilisers stabilise incomes, not the money supply. The fact that taxes are cleared by money payments is incidental; what matters is that government liabilities are drained.

      Delete
    3. Brian, perhaps talking about the nature of money is a tactical mistake when talking to normal people. For me, coming from an economics background, it was the sense of logic that MMT demonstrated and it made complete sense and showed to me just how many of the foundations of the economics I had learned were wrong. It destroyed the claims of what I had been taught, at least for me.

      As for your point #2- are you are saying economics should not really concern itself with definitions of the money supply, and therefore should abolish the various amounts of the different 'money supplies' from any primary consideration? I could agree with that- 'the money supply' in economics is not the 'money' that people and firms produce for in a 'monetary economy'.

      Since I am ornery today, yes the automatic stabilizers stabilize incomes and not the money supply. I didn't mean to say that they stabilize the money supply (if I did). But for most people, isn't 'income' practically the same as what 'money' they will have for their spending in the near future? And that spending of 'money' is what most often convinces the people at the electric company to continue providing electricity, and the grocery store to give them food, and so on?



      Delete
    4. Ralph Musgrave, I agree most people dislike paying taxes, but most people who provide services to the government really, really hate when those are commandeered from them and are much happier to be paid with money for them. In my opinion.

      And a problem of governments just taking, rather than paying people to provide, is that they can only take what is already available in the form of goods and services. And I suspect that what that government starts commandeering will soon decrease in supply available. I mean who is going to plant corn if they know the government is going to confiscate corn and the only people they can confiscate corn from are people with corn at that time?

      Delete
  9. “Let us assume that all workers are paid a uniform wage, whether they are working for the private sector, or enrolled in the Job Guarantee.” Nothing wrong with that assumption: I always believe in solving over-simple and hypothetical instances of a problem before moving on to the real world.

    However there is a big problem there which about 95% of JG advocates never admit to, which is that if people can get as much on a JG job as they can on a normal or regular job, why should they bother job searching? Put another way, generous JG pay reduces aggregate labour supply, which is inflationary, which in turn means aggregate demand must be reduced, which destroys regular jobs: hardly the object of the exercise.

    The Swedish labour market economist Lars Calmfors concluded many years ago that the only solution to that is relatively low JG pay.

    Alternatively, people could be shifted from productive regular jobs to working for state employment agencies which find work for JG people and then talk or persuade JG people into shifting to those regular jobs. Or perhaps JG people could be trained to do that employment agency work. However, there are ten thousand different jobs out there: e.g. there are dozens of different types of plumbing or electrical work. Only those in possession of relevant skills really know what they are suited for: employment agency staff cannot have the same knowledge. Thus I suspect (like Calmfors) that relatively low JG pay is the only solution, combined with some sort of Workfare type sanction: i.e. “do this JG job else your unemployment benefit gets reduced”

    ReplyDelete
    Replies
    1. It’s a hypothetical example, AS I HiGHLIGHTED IN THE TEXT.

      Every serious discussion of the Job Guarantee I have seen notes that it becomes the de facto minimum wage for the eocnomy. You do not need a doctorate in economics to figure out that there will be an impact on employers who rely on minimum wage employees. They will be forced to either pay relatively more (creating a flatter income distribution within the private sector), offer non-wage benefits, or make working conditions more attractive. Other than for the cheerleaders for bottom feeding employers, these are actually attractive effects.

      Delete

Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.