I'm optimistic about US recovery. If I'm reading the signs right, the market is also optimistic about US recovery. But that's not what makes me optimistic. Again, if I'm reading the signs right, the market is more optimistic about US recovery than the Fed is. And the market believes it is more optimistic about US recovery than the Fed is. That's what makes me optimistic about US recovery.
Let's start with a bit of Wicksell. If the Fed sets a market interest rate above/below some natural rate, there will be a cumulative decline/rise in the price level. Throw in a bit of Keynes. And there will also be a decline/rise in real output too.
Now let's throw in some expectations. The actual natural rate matters. But what the market believes the natural rate is matters too, and probably matters even more than the actual natural rate. Start in equilibrium. Hold the actual natural rate constant. Hold what the Fed does constant. Now suppose the market changes its beliefs and (falsely) believes that the natural rate has increased. And suppose the market also believes that the Fed does not share the market's view, so will not change what it does. And suppose the market understands Wicksell, and Keynes. What happens?
Since the actual natural rate has not changed, and since the market rate set by the Fed has not changed, you might think that no individual will want to change his desired savings or investment. But, because each individual (falsely) thinks that the natural rate has increased, relative to the market rate, each individual thinks that every other individual will increase desired investment and reduce desired saving. So each individual expects the Wicksellian/Keynesian cumulative process will cause rising prices and output. And this is what causes each individual to increase his own desired investment and reduce his own desired saving. And so there is a Wicksellian/Keynesian cumulative process of rising prices and output.
Somewhere, deep in the metaphysical vaults of preferences and technology, there exists a true natural rate. And if everybody knew what it was, and the Fed did too, and set the market rate equal to it, the economy would be in metastable equilibrium, with no cumulative process in either direction. But the true natural rate matters much less than what the market believes the natural rate is. Because even if you are impatient, or have a good investment opportunity, you will not consume or invest as much if you expect falling prices and falling spending by everyone else.
And beliefs about the future matter too. If the market believes the Fed will set a market rate below the natural rate in future, then the market will expect a Wicksellian/Keynesian cumulative rise in prices and output in the future. And that causes a rise in desired investment and a fall in desired savings today. And so the cumulative process starts today, not in the future. And the Fed would have to raise the market rate right now, to prevent it.
It's the gap that matters. In the simplest Wicksellian story, the gap is the gap between the natural rate and the market rate. Throw in expectations about the future, and what matters is the gap between what the market believes will happen and what the market believes the Fed believes will happen. I think we have just such a gap right now. The market believes the Fed is too pessimistic. That creates an upside cumulative process. That's what makes me optimistic.
I wonder if the Fed might be deliberately making pessimistic noises (Tim Duy, via Mark Thoma) just to make the market optimistic, like me?
Worthwhile Canadian Initiative 
JKH: Oh God. Since you read their stuff, and are a careful reader, you are probably right.
I had previously understood MMT as a cyclical doctrine.
“Cyclical MMT”: in recessions use money-financed fiscal deficits to increase AD, and in booms use money-financed fiscal surpluses (i.e. reduce the money supply) to reduce AD. Not obviously bad policy, though it may be difficult to change fiscal policy quickly, and there may be microeconomic reasons why you don’t want to keep changing tax rates and government spending like that. Cyclical MMT is not obviously stupid.
But what you are describing is “structural MMT”. One great big and permanent QE2, regardless of whether we are in recession or boom. And that idea is catching on like wildfire on MMT blogs? If so, I take back anything nice I ever said about them (except Warren Mosler makes nice cars). God only knows where to begin. I’m not so sure I can engage them in discussion.
Even Abba Lerner would reject structural MMT.
Nick, I think you’ve misunderstood me. I’ll be back with additional interpretation, hopefully clearer.
JKH: “From an MMT operational perspective, people in aggregate have little choice.
The government deficit spends by handing you a check. You deposit in your bank account. Reserves increase.”
Of course. But that’s just (sorry) the accounting. Monetarists (for example) agree with that. But if people don’t want to hold $8 trillion in 0% chequing accounts, they will try to spend it. And they will bid up the price of everything else as they try to get rid of it. And prices will rise until prices get high enough that they are willing to hold all of that $8 trillion and stop trying to get rid of it. And if they were only willing to hold (say) $1 trillion at 0% at the existing price level, that means, to a first approximation, you need an 8-fold rise in the price level. 700% inflation for one year, say.
OK, it’s not quite Zimbabwe, but it’s getting close.
Nick,
“But what you are describing is “structural MMT”. One great big and permanent QE2, regardless of whether we are in recession or boom.”
No.
What I’ve tried to describe most recently is structural change in the FORM of deficit financing – not change necessarily in the mix of structural and cyclical deficits, which is a separate conceptual issue using the same vocabulary.
With respect to FORM, instead of issuing bonds, the government allows all net disbursements (deficits) to settle as increase deposits and increased reserves in the banking system.
The “zero natural rate” idea is consistent with the idea that the government/central bank pays no interest on those reserves.
Such a structural form change is an option under MMT, but heavily favoured by Mosler and Wray.
My reference to the $ 8 trillion number in the case of the US can be interpreted in the following way:
Had the MMT “no bonds” proposal been operational already, starting back in the past, the entire deficit would have been “funded” with bank deposits/reserves and currency. Specifically, outstanding bonds would have been replaced by bank deposits and corresponding excess reserves.
In contrast with that, the size of deficits is an entirely separate issue. It’s definitely an issue, but it’s different than the issue of what financial system configuration accommodates those deficits.
I’m using the word “structure” in one way here in the sense of the financial architecture or the financial structure or the macro balance sheet configuration under which the government will run its deficits. I am not referring here to the difference between structural and cyclical deficits.
My earlier comment however was to suggest that this change in financial architecture, MMT’ers argue, makes it easier to “run deficits” in the sense that government financing is not held hostage to bond market hysteria and bond market “vigilantes”, because bonds are no longer issued. I also alluded to the fact that my own interpretation of MMT policy orientation is that it favours if anything an approach that almost always includes deficits – there is no necessary idea of deficits in one environment offset by surpluses in another. They also have views on the issues of structural versus cyclical, which I don’t feel I can comment on here accurately.
Hopefully that’s at least epsilon clearer. Sorry if I confused. No doubt you can translate this into vocabulary that is less ambiguous.
“Of course. But that’s just (sorry) the accounting. Monetarists (for example) agree with that. But if people don’t want to hold $8 trillion in 0% chequing accounts, they will try to spend it. And they will bid up the price of everything else as they try to get rid of it. And prices will rise until prices get high enough that they are willing to hold all of that $8 trillion and stop trying to get rid of it. And if they were only willing to hold (say) $1 trillion at 0% at the existing price level, that means, to a first approximation, you need an 8-fold rise in the price level. 700% inflation for one year, say.”
I understand that argument.
But this structural (sorry) information about the “no bonds” option that I’ve apparently revealed to you here gives you additional context for meeting any argument about inflation transmission that MMT’ers might care to present in debate – should you choose to reconvene that debate.
I actually recommended to them that they develop an exposition of how they view their own theory of inflation and elevate it to a much higher level in their public presentation – precisely for the reason that is implicit in you presenting your case. I think their structural option of “no bonds” makes that presentation essential to their overall case. Sadly, I got no response to that particular comment. They’ve certainly written about it, but in my view it needs a higher profile in their “theory” content (as opposed to their “accounting reality” content) within MMT.
Nick,
On an entirely separate subject, I’d be interested in your thoughts on my comment/question at Sumner’s Delong reply post, if you get a chance sometime. Some economics that I probably don’t understand (what a surprise):
http://www.themoneyillusion.com/?p=8136#comment-48352
JKH: Yes, that’s clearer. That’s what I call “cyclical MMT”, but with the added proviso that cyclical MMT had always been in place, since the year dot, when governments first started.
What it does is takes away one degree of freedom from government policy. For example, if there’s some temporary need for increased government spending (war, floods, new schools for a baby bulge, etc.) and the level of aggregate demand is currently at the desired level, so you don’t want to use money-finance, then under MMT the government must raise current taxes to keep a balanced budget, rather than spread out the higher taxes as under bond-finance.
It’s unwise, because tax rates would be more volatile, so the deadweight costs of taxation would be higher, and also difficult to implement, given lags in changing fiscal policy, but otherwise OK.
On your last point. Yep. They definitely avoid the topic of the LRAS or Long Run Phillips Curve. I tried to draw them out on this about a year ago, but no luck.
Dead weight losses might be higher or lower. Commercial bank seignorage is a huge dead weight loss. So is the existence of income or consumption tax when land value isn’t fully taxed. I’m still not certain that MMT isn’t at least in part equivalent to full land tax plus government seignorage only, neither of which I consider inefficient.
JKH:
“I actually recommended to them that they develop an exposition of how they view their own theory of inflation and elevate it to a much higher level in their public presentation – precisely for the reason that is implicit in you presenting your case”
What, you mean something as coherent and demonstrable as “Inflations expectations theory”? Winterspeak already adequately exposed the weakness of the inflation expectations paradigm. I know you’ve been on Mitchells blog frequently. Ive seen you comment a lot. You’re quite aware of how they deal with inflation. Just because the average person has no fricking idea about deflation/inflation, believes ZImbabwe and Weimar were just govts printing money and feels better when a monetarist says “we’ll just raise rates if we sniff inflation” doesnt mean the subject hasnt been adequately dealt with by the MMT crowd.
Yes the non MMT crowd talk endlessly about inflation. They see it everywhere and obsess continually about it, but they have been horribly wrong in most every prediction made the last decade.
Gizzard,
You seemed to have missed my point, which is not that MMT should kowtow to a mainstream inflation theory, but that they should elevate their communication of why they don’t kowtow to it, including more elaboration on their central idea of “real constraints”. My point refers to communication emphasis, in the context of the full MMT message, not whether they’ve developed their argument per se.
Adam P,
The particular problem I was thinking of has to do with longer maturity bonds that pay coupons. There is no “theory-free” way to strip the coupons in real terms, and obtain a single “real” payout when the bond matures. You need some assumptions about the relationships between shorter term rates and longer term rates, as well as re-investment risk. That is what I meant by EH.
Now perhaps you don’t care about that, and you are willing to say that a coupon paying bond over n periods is a vector of (n+1) cash-flows, each of which can be divided by the price level at the time the cash-flow is obtained. That would be a an “objective” measurable value in the underlying economy.
But in that case, you have an nxn vector of “real” interest rates, just for coupon paying bonds. Then add in zero coupon bonds and more complicated instruments, etc.
Now in the models, investment decisions are made based on longer term rates, not single period rates. I would argue that savings decisions are also made by the longer term rates. So to the degree that the economy is influenced by real rates, you would have this problem.
Then you have the additional issues of how you would price such a bond given heterogenous preferences — i.e. can you explain the rate in terms of preferences and endowments. But there you can assume some pricing kernel if you want an EMH framework.
The point is that the cash-flow inputs into that framework need to be directly observable, or known, apart from the framework itself, and then you use people’s own utility and beliefs about the likelihood of the state of the world to value the cash-flows. But the cash-flows themselves are nominal, not real; there is no notion of a “real cash-flow” that is not itself model dependent — as far as I know.
So then the question becomes, if “real rates” are model dependent, then in what sense does the economy have a real rate? Only in the sense that the models converge to accurately describing reality. This is a very strange condition to impose on a price. Prices are directly observable in the economy and exist outside of the model. The model is there to explain the prices.
In the above, I meant to say “in the economy, investment decisions are made by longer term rates”, not “in the model” 🙂
“will people want to hold all $8 trillion at a 0% nominal interest rate?”
From an MMT operational perspective, people in aggregate have little choice.”
But of course, people do have a choice. So the government would need to take that choice away. In the current framework, banks must compete for deposits and they must obtain funding from the private sector. This is the mechanism that allows households to decrease their deposit holdings and increase their holdings of financial sector bonds, for example.
One can imagine a radically different regulatory environment in which the government allows banks to obtain funding directly from the government rather than from the private sector, and also engages in price-fixing to prevent banks from offering competitive certificates of deposit, etc.
But even then, the government would need to ban money market mutual funds and other institutions that offer longer duration liabilities.
In that case, the resulting framework is one of bank oligopolies and financial repression, which I find very troubling.
Real world examples of such frameworks have not had positive outcomes.
When the government forces interest rates to be excessively low by outlawing or destroying capital markets, the result need not be excessive consumer price inflation, it could be excessive investment and asset price inflation, as well as growing income inequality, because by lowering the rate of discount, you are increasing the endowments of some actors (e.g. those with capital), and unless everyone holds the same amount of capital, this will result in large levels of inequality.
The underlying MMT economic framework assumes that investment and savings are completely interest inelastic, and moreover that the profit rate is equal to the rate of interest by some magic law, rather than by a process of competitively pricing interest rates against equity returns. When interest rates are not competitively determined, the only mechanism to bring down the profit rate to the interest rate is excess capital investment — e.g. dynamic inefficiency. The outcome of such a policy is going to be falling living standards and growing inequality.
RSJ: “The particular problem I was thinking of has to do with longer maturity bonds that pay coupons. There is no “theory-free” way to strip the coupons in real terms, and obtain a single “real” payout when the bond matures. You need some assumptions about the relationships between shorter term rates and longer term rates, as well as re-investment risk. That is what I meant by EH. ”
What your talking about has, again, nothing at all to do with the conceptual problem of defining a real rate of interest. If you have a (coupon) bond with maturity equal to each coupon date then one can bootstrap out a zero coupon curve in a theory free way. If short-selling is allowed then we can even trade the zeros.
But all that is any completely irrelevant to talking about real rates as a theoretical construcct. Everything I said before still stands, the difficulty is getting everyone to agree on the price index that defines the price of their consumption basket. If you could do that then in there will exist a shadow price of the real zero coupon bond.
This definition is in absolutely no way dependent on any model. If today you introduced a new maturity of nominal coupoon paying bond I’m very sure the market would find prices to trade it at. I’m sure of this because it happens all the time.
If today you introduced a zero coupon nominal bond that isn’t already trading, I’m sure the market would find the price to trade it at.
If today you introduced a zero coupon real bond, (subject to everyone agreeing it actually is a real bond), then I’m equally sure the market would find prices to trade it at.
This absolutely nothing to do with any model, theory or assumptions about the EH.
As I said above, the substantive problem with trying to define a real rate have to do with defining the consumption basket that everyone agrees on. What you’re talking about has absolutely nothing to do with anything.
You can’t convert one cash-flow to another in a theory free way.
You can “construct” zero coupon versions of market traded bonds, using theory, but your constructed zero coupon bonds are fictional, and won’t trade for the same price as a real zero coupon bond in the marketplace, primarily because stripping assumes that the yield curve is not going to change throughout the life of the bond — you are assuming that market prices are martingales, which is a big theoretical assumption.
Just to save time, any process that can be described as “valuation” requires theory.
The only theory free thing you can do is make a market for the two cash-flows and watch the relative nominal prices bounce around in the marketplace.
But that indifference price will be a function not just of the two specific cash-flows, but of everything else — economic outlook, risk aversion, wealth levels, preferences, different market participants, etc — i.e. it will be time-varying. And you can’t untangle these effects without theory.
The market prices that we observe — at least in a monetary economy — are nominal prices. Just because everything has a price does not mean that everything has a “real” price.
Only real goods have real prices.
If you cannot objectively measure and/or quantify the good, then you cannot assign a real price to it unless you rely on theory.
I.e., if it costs $5 to purchase a loaf of bread, and $10 to purchase a pound of cheese, we can talk about the price of bread in terms of cheese, i.e. “real” prices.
And we can say that next period, something in the economy changed so that the real price of bread (in terms of cheese) went up.
But implicit in that statement is that you found an identical slice of bread and an identical slice of cheese, and compared the prices.
There needs to be some objective, or price-independent, definition of the bread and cheese, so that you know that you are comparing identical goods from period 1 to period 2.
Now what does it mean to say that the “real” interest rate went up, or went down, from period 1 to period 2? What is your price-independent method of finding an identical bond to compare against the identical slice of cheese? If the economy changes, then you cannot find an identical bond in period 2. You need some assumption — that risk premia are constant, or that preferences for cheese are constant, or that preferences for bonds are constant, etc.
RSJ,
“You can ‘construct’ zero coupon versions of market traded bonds, using theory”
Yes. You can also construct them in practice. US treasuries, Canadas, Bunds, Gilts, whatever. Just strip the bonds. People do it all the time and in practice they trade where predicted by theory: At a slight, but for macroeconomic purposes, insignificant discount to the on-the-run curve, for the simple reason that they are not repo special. Basically they trade right on top of the off-the-run government bond curve.
But even if you don’t want to strip bonds you can hedge out the “reinvestment risk” by constructing an appropriate portfolio (long and short) of regular treasuries. There is just no issue in constructing zero coupon bonds, theoretical or practical.
“stripping assumes that the yield curve is not going to change throughout the life of the bond”
No it doesn’t. It assumes you can strip the bond. Which you can.
“Just to save time, any process that can be described as “valuation” requires theory.”
If by “theory” you mean interpolation. And whatever interpolation method you use won’t make any difference to the zero yield at least for macroeconomic purposes (i.e. within a few basis points).
“you are assuming that market prices are martingales, which is a big theoretical assumption.”
No he isn’t. He’s probably assuming the law of one price. But even the EMH does not assume any market to be martingale. The whole market is usually thought to be a sub-martingale, but in the absence of a risk-free asset, even this is not implied.
K,
I’m not saying that it is difficult in practice to engineer zero coupon bonds out of other bonds. I’m saying that until you let the market price the bond, then any predictions of what the bond price will be are going to be theory-dependent. This is completely independent of whether those predictions tend to be correct or not. It may be that the theory is not too far off.
The “Law of one price” is violated frequently and prominently. Google “Liquidity, Reconstitution, and the Value of U.S. Treasury Strips”, but there are many other examples. Now you will say, as do the authors, that this is due to some liquidity premium, which is the whole point, right?
Market prices — the observables — are a combination of the cash-flows and preferences. And market prices are the only way to aggregate the cash-flows. Therefore all you can say is that cash-flows + preferences = price. You cannot say cash+flow + objective algorithm = price. The algorithm will need to take preferences into account — it will be model-dependent.
RSJ,
As K has already pointed out this, “stripping assumes that the yield curve is not going to change throughout the life of the bond — you are assuming that market prices are martingales, which is a big theoretical assumption.” is utter nonesense.
Anyway, I can’t really make sense of anything you’re saying here and can’t see how it’s relevant to the original debate so I’ll just leave it at that.
We can both remain content with the correctness of our view.
K
re: assuming the curve being constant — yep, I was totally wrong about that. I kept thinking about re-investment risk, but that’s irrelevant here.
Adam,
Sorry, I didn’t see your message. Yep — wrong about the curve being constant. If you don’t understand what I am saying about theory-depence, I’ll try one last time. Perhaps I am totally wrong there, too?
The claim is that the nominal rates are what are observed in the economy, whereas the real rates are inferred by the model, and are model dependent.
OK, say you define the real rate over 1 period as the nominal rate – expected inflation over that period. Say you can observe expected inflation (e.g. by a poll of investors). So you have nominal rate = real rate + expected inflation.
Then you will still end up with discrepancies, in the sense that two different riskless instruments that mature in the same period will have different prices. So you attribute that to a liquidity premium term. Now you have
nominal rate = real rate + expected inflation + liquidity premium.
And then you do some more measurements, and you determine that even instruments that are equally liquid, but perhaps have somewhat different tax-treatments, also have different prices. So now you have
nominal rate = real rate + expected inflation + liquidity premium + tax treatment
Etc.
So the claim is that the nominal rate is the only “real” thing. And the rest of the stuff is model dependent. In the real economy, there are many different bonds and many different interest rates. And attempts to determine a consistent decomposition into more fundamental terms isn’t going to be consistent, at least to the degree that you can isolate and independently observe all the constituent factors. All you know is that asset demand will set the price.
OK, that may be vague, but is hopefully sensible.
On my way out of town, but got a Google alert on this, so thought I’d throw in a few points, mostly for clarification of the MMT view for those that are interested (rather than as a critique of anyone commenting here). Apologies for not being able to engage further, but it’s off to the in-laws early in the morning. Hope all is well, Nick. Long time, no chat!
1. The discussion of the “natural” rate in the MMT view–Mosler/Forstater’s paper (the seminal paper on this) is very clear that the “natural” rate in MMT is not the same thing as the “natural” rate that Nick is speaking of, as JKH noted. One thing missed in the discussion here that is crucial to Mosler/Forstater is that the non-govt sector on average will prefer to net save, given that they are currency users, not currency issuers. Historically, aside from the 1998-2008 period, this has been true (this is partly why MMT’ers were so alarmist about building financial fragility during this period). In that case, aside from a sizeable trade surplus, this will require govt deficits, on average (again, the 1998-2008 period is an exception well-understood by MMT’ers). As such, it is deficits that are “natural” in the argument more so than the zero rate–the “zero rate” is simply along for the ride if the govt doesn’t sell bonds or pay interest on reserves. Then there’s the policy side advocating the zero rate, but for totally separate reasons. (As an aside, I’m not personally in favor of the “no bonds” proposal that most other MMTer’s advocate.)
2. On the real rate, as Nick suggested, MMT’ers reject that real rates matter, as Keynes also did. Jan Kregel has explained this very carefully many times. Eric Tymoigne has done a bit on this recently. There’s also a good deal of empirical evidence, at least persuasive to MMT’ers.
3. Regarding inflation, the reason you don’t see too much MMT stuff on the specifics of an inflation theory is because in general the PK view is accepted and there’s already a good deal of literature there. The PK view relies on a number of things–bottlenecks, pricing power, markups, commodities, bargaining power of labor–but it does generally reject overly strong reliance on inflation expectations, a la Winterspeak. Bill has done a lot of empirical work on unemployment/inflation tradeoffs–he’s an econometrician by training. Lavoie/Kreisler have done some work in which they cited several empirical Fed papers on inflation/unemployment that are completely in line with the PK view.
Best,
Scott
Thanks Scott:
I like to try and make sense of various views of economics, even if I don’t agree with them.
“One thing missed in the discussion here that is crucial to Mosler/Forstater is that the non-govt sector on average will prefer to net save, given that they are currency users, not currency issuers.”
That’s the bit that’s least clear to me. My guess is that “net save” means something like “savings minus investment”, and that what MMTers are saying here is something akin to what I would say this way: “In a growing economy, the desired stock of real currency is growing over time, and the government must satisfy that demand by issuing currency in order to avoid deflation”.
On the Phillips Curve: a number of people have noted that the Phillips Curve has disappeared, or “gone flat” over the last 20 years or so. My interpretation is that this is what we should expect in any country where the central bank is targeting inflation reasonably successfully. All fluctuations in inflation should become unforecastable noise, because if any fluctuation were forecastable, the central bank would have done something different to have stopped that fluctuation.
On the vertical IS curve (perfectly interest-inelastic savings and investment): this is the assumption that will bother other economists the most. Because it’s like saying (in an intertempoarl context) that relative prices don’t affect demand. Demand curves are vertical. Also, if the IS were vertical, yet the Bank of Canada thought it sloped down, we would have seen very large (infinite?) fluctuations in real interest rates. If the Bank of Canada wanted to reduce demand it would raise interest rates, and when it saw it wasn’t working it would raise them even more…and so on.
Scott,
I haven’t been exposed to MMT per se for that long, but it’s been long enough to have developed a view as to which points might be emphasized if you and your colleagues really want to get the message out more broadly. Quite apart from the ELR idea, I’d recommend central focus on two main ideas in connection with the purely monetary side of things:
1. Position the “no bonds” IDEA prominently in the main message. I say “idea” rather than “proposal”. My view is that as soon as you begin to touch on the notion of “self-imposed constraints”, you simply can’t avoid referencing the conceptual apparatus that associated with a “no bonds” hypothetical. Whether or not it is an actual proposal favoured by some or not is secondary.
2. Notwithstanding your point regarding inflation, the MMT view of it should still be emphasized. Combined with the idea of constraints, it is essential to the explanation of the MMT approach to deficits. I came to this view about the central importance of the inflation story after viewing the discussions at the link below, which include a number of very interesting conversations about inflation (e.g. Mosler) – that almost emerge as side-stories instead of the main deal (and I’m not referring to Auerback’s hyperinflation presentation):
http://neweconomicperspectives.blogspot.com/2010/10/fiscal-sustainability-teach-in-and.html
This is not to say that these ideas haven’t been covered in great depth already. Obviously they have. I’m just saying, given the not unnoticeable frustration of some MMT’ers that much of the world seems immune to absorbing their message (a frustration that I think was expressed strongly at the conference), I would recommend that these are the two points that should be given prominence in a popular teaching framework. In that regard, I was disappointed there was no response to my comment at the site. And along with that, I have to say that my general impression is also that MMT’ers aren’t all that receptive to constructive criticism about the way in which they present their message. Of course, my recommendations may just be lousy, but I think I understand at least a bit about the underlying monetary concepts.
Nick, should you have some time over the holidays, I recommend sampling these discussions. They’re very high quality in terms of substantive MMT content and differentiation.
JKH: For someone like me, if the MMTers would just write down a simple model, it would be a lot more useful. I would understand that it’s a simple model, and like all simple models, would leave stuff out. That’s OK. But what it would contain would be the main elements that MMTers think are most important. And it would also be helpful if they were very clear on the distinction between real and nominal variables. I keep having to guess at what they mean.
So far, my guess is that their model has a vertical IS curve, a horizontal LM curve, and a reverse-L-shaped AS curve or Phillips Curve. Essentially, it’s the old British Keynesian model from the 1960’s. It’s what I learned as an undergrad, only with a lot of strange terminology and more accounting, none of which really changes the essentials of the model, but does make it harder to understand.
OK, time for one last one this morning–I’m a bit of an addict, so even when I know I’m going to get in trouble for making us late, I still gotta sometimes.
Anyway, all good comments.
On inflation, JKH, I don’t disagree at all. The issue is two-fold, perhaps (can’t say exactly), for why it’s not emphasized. First, inflation isn’t the issue right now, though, of course, it’s certainly what bothers a lot of our critics. Second, we’re only a few handfuls of economists working on MMT. We’ve done a lot, but we haven’t done everything. Our main contributions on framework building, in my view, are to explain how the monetary system works, elaboration and application of the Minskian view of fragility, and Mitchell’s very innovative work on labor markets. Beyond those, we’ve “borrowed” the rest, mostly.
On “no bonds,” I’ve definitely explained a few places the reserves/deposits vs. money issue. Yes, “no bonds” may/may not be front and center, but reserves/deposits vs. bonds is. Again, though, complete framework incorporating inflation and such, maybe not (though we would argue it’s not necessary for our purposes).
On Nick’s points regarding model, IS-LM won’t work for us. There isn’t 1 interest rate, there are many, and all of them can “matter” at different times, and for various reasons. IS’s elasticity isn’t fixed. The point about the real rate is that it isn’t necessarily the “real” rate that matters, but other things do–mortgage rates can matter, the overnight rate can matter but might not at all, credit default spreads definitely can matter, cost of capital can matter (but is often swamped by expected cash flow) and so forth. As for LM, same thing–it’s way too simple to capture what we’re talking about. Yes, banks respond endogenously to demand for credit, but there are issues like capital requirements, and various shift factors (again, credit default spreads, etc.). As for Phillips, roughly horizontal, but not completely, and with many shift factors, again, and we would argue this works for both pre- and post-inflation targeting by CBs.
Best,
Scott
Two solitudes in a way.
(Being Canadian helps to appreciate that reference.)
Somewhere, there’s gotta be a concise story about MMT that is a blend of an unusually intense operational focus, combined with another part that is a borrowed/filtered/reworked theory about inflation, etc.
Perhaps the operational side could be told with some assistance by adjusting the lyrics to John Lennon’s “Imagine” with the line “Imagine no bonds” thrown in there. It’s timely.
Let me throw this out there:
Question. What is they key price or quantity that needs to adjust to ensure macroeconomic equilibrium?
A1. (New Keynesian): The real rate of interest.
A2. (Monetarist): The real stock of money.
A3. (Old British Keynesian): The government budget deficit or surplus (however financed).
A4. (MMT): The (money-financed) government budget deficit or surplus. ?
Nick, yes, it would be A4, though for MMT there is no other type of deficit besides “money financed.” In other words, “however financed” for us misses the fact that they are all equivalent, assuming flexible fx and currency issuer. The MMT understanding of the monetary system is very different from that in A3, and A2 for that matter (though MMT’s understanding is actually closer to A2 than to the others). A1 (assuming “new consensus” here) doesn’t really have a monetary system (or at least not a financial system), just an interest rate.
Scott: Right. I’m glad I’m more or less getting it.
1. In the MMT perspective, if the central bank “does nothing” then government deficits are automatically money-financed.
2. Ignoring 1, one main difference between monetarists and MMTers is this: monetarists say it’s the actual stock of (real) money that has to adjust to the desired stock; MMTers say it’s the actual flow of new (real) money that has to adjust to the desired flow. This is not just a semantic or modelling difference. If you took the derivative of the monetarist’s stock demand for money function, you would not get something that looked like the MMT flow demand for new money function.
3. Ignoring 1 and 2, when it comes to the flatt(ish) Phillips Curve, and the vertical(ish) IS curve (interest-elasticity pessimism), MMTers are closer to Old British Keynesians than anyone else.
Scott: “A1 (assuming “new consensus” here) doesn’t really have a monetary system (or at least not a financial system), just an interest rate.”
Yep, assuming “new consensus” here, New Keynesians think A1 doesn’t have a monetary system, just an interest rate, but it does really, though it’s implicit, so they can’t see it. I seem to be in a minority in arguing that NK makes no sense in a barter economy. I take it that MMTers agree that MMT would be total rubbish in a barter economy. If (hypothetically) barter did break out, they would throw away MMT and start from scratch.
What if a major bank collapses once Wikileaks exposes their true financial situation, and people start asking what are other banks still hiding on their books? No one on this blog seems to realize that the heart of the current crisis is massive financial fraud and the ongoing coverup by corrupt governments and central bankers.
I would also like to see a simple (non-structural) model. If many interest rates are important, then just start with two rates, and you can describe the interaction. To my mind, there is nothing that prevents MMT from doing this.
I don’t think Marc Lavoie is MMT, but he has a model with Wynne Godley,that I sometimes see referenced. May make explicit the old british keynesian/post keynes relationship, I think Godley worked at the uk central bank, back in the day, and was motivated by that to do his work on accounting.
edeast
Lavoie isn’t necessarily MMT, but there’s not much if anything that he and MMT disagree upon. MMT is strongly evidenced by the Godley/SFC approach and also by horizontalism. Lavoie’s understanding of monetary operations (treasury, CB, banks) is essentially identical, as well, to MMT. The only difference I’ve noticed over the past year is that Lavoie and his colleagues don’t find much use for Minsky in the recent financial crisis, whereas Minsky is front and center for MMT.
Nick,
1. Yes, MMT wouldn’t apply to barter economy.
2. Regarding, NK/NC and whether they have a financial system or not, I’m referring here to, say, Buiter’s critique, but also more importantly to Goodhart’s critique of DSGE models (granted, one can be NK/NC without having a preference for DSGE) or the Woodford-type models.
Scott if you’re still reading I’d love to see you address this, quite correct, point of Nick’s:
“On the vertical IS curve (perfectly interest-inelastic savings and investment): this is the assumption that will bother other economists the most. Because it’s like saying (in an intertempoarl context) that relative prices don’t affect demand. Demand curves are vertical. Also, if the IS were vertical, yet the Bank of Canada thought it sloped down, we would have seen very large (infinite?) fluctuations in real interest rates. If the Bank of Canada wanted to reduce demand it would raise interest rates, and when it saw it wasn’t working it would raise them even more…and so on.”
Do MMTers think demand curves are vertical, that relative prices have no effect on relative demands?
Of course I’d also be interested to hear any other MMTer (say JKH or Winterspeak) answer…
I’m not an MMTer, but I could list a few things other than rate/price that could influence, or put a limit on demand at the individual level.
Level of current income, level of current expenses, current level of indebtedness, capacity to service existing debt, capacity to borrow more-to financing more spending, confidence in stability of future income, expectations about future increases in fixed expenses, confidence in future value of purchasing power of current savings (if have any), non-existence of any lender calling on their debt, confidence in future positive value of existing savings/investments/net worth, ease of liquidating current savings –to finance current spending, tax implications of using current income/net worth to spend now vs. later, confidence about over-all economy, and how it is being run… the list goes on further.
If the challenge were to simply reduce demand, increasing prices/rates will do. But if the challenge were to increase demand, you also have to address the above
I think Scott’s point was not that the IS curve is vertical — he didn’t argue that — but that there is no such thing as a single economy-wide IS-curve because there are multiple interest rates that matter, and the relative importance of these rates is situation dependent.
Moreover, a lowering of one rate might result in the increasing of another rate. For example, if overnight borrowing becomes cheaper, then inventory financing becomes cheaper and dividend yields may go up, not down. Or at least, there is no reason to believe that dividend yields will fall if overnight rates fall.
There is no micro reason for this, nor is this relationship observed, nor does it described any sort of plausible constraint on investor behavior. It’s an arbitrary constraint plucked from thin air and imposed on the economy.
And Scott also argued that the interest elasticity of investment was also varying.
There are no micro-foundations for an economy-wide IS curve, and it may be the wrong constraint to put into the model.
No one is arguing that relative prices have no importance.
RSJ and rogue,
I’m not asking about the IS-LM model. I’m asking about this quote from Scott Fullwiler: “On the real rate, as Nick suggested, MMT’ers reject that real rates matter,”
Which exactly contradicts RSJ’s: “No one is arguing that relative prices have no importance.”
Scott is arguing that relative prices have no importance along certain dimensions.
I’m looking for some sort of defense of that position.
Adam “I’m looking for some sort of defense of that position.”
My guess is that the defence will involve income (wealth) effects offsetting substitution effects.
My own position is that in macro, to a first approximation, unless there are good a priori reasons to suppose otherwise it’s the income effects themselves that (roughly) cancel out. When interest rates fall, creditors are worse off, and debtors better off, so at the macroeconomic level what’s left over is a distribution effect, not an income effect. Plus the substitution effect, of course.
Like Old British Keynesianism, substitution effects are downplayed, and income effects are upplayed in MMT. My guess also fits with the MMT emphasis on accounting. Accounting tells you nothing about substitution effects, but it can be useful to help you sort out income effects.
Nick, Adam P,
“My guess is that the defence will involve income (wealth) effects offsetting substitution effects.”
That rang a bell.
See:
http://bilbo.economicoutlook.net/blog/?p=12473
“The worker is conceived of at all times making very complicated calculations – which are described by the mainstream economists as setting the “marginal rate of substitution between consumption and leisure equals to the real wage”. This means that the worker is alleged to have a coherent hour by hour schedule calibrating how much dissatisfaction he/she gets from working and how much satisfaction (utility) he/she gets from not working (enjoying leisure). The real wage is the vehicle to render these two competing uses of time compatible at a work allocation where the worker maximises satisfaction.
What happens when the relative price between C and L changes? The final result looks scientific but is in actual fact a total fudge.
They isolate two separate “effects” of such a real wage change (say a rise): (a) a substitution effect; and (b) an income effect.
So an increase in the real wage (more corn is foregone for an hour of leisure) leads the worker to consume less leisure and to work more. This is the substitution effect. So if the real wage rises, work becomes relatively cheaper (compared to leisure) and the mainstream theory asserts via the so-called law of demand that people demand less of a good when its relative price rises. So real wage up, less leisure, more work.
But there is another effect to consider – the so-called income effect. When the real wage rises, the worker now has more income for a given number of hours of work.
The mainstream theory of normal goods (as opposed to inferior goods – the distinction is just made up largely) tells us that when income rises a consumer will consume more of all normal goods. The opposite is the case for an inferior good.
Leisure is considered to be a normal good as are other consumption goods the worker might buy with the income he/she earns. So as the real wage rises, the income effect suggests that the worker will demand more of all normal goods (because they have higher incomes for a given number of working hours) including leisure. That is the worker will work less and consume more leisure.
What is the net result? No analytical solution is provided. They just assert that the relative price (substitution) effect is stronger than the income effect and so the labour supply curve is upward sloping. It is totally made up result and no robust empirical analysis has supported this assertion. The reason they need to assert this result is because they also assert the demand curve is downward sloping and for an “equilibrium” they need the curves to cross. Simple as that.”
JKH, I asked about real interest rates.
I know.
Just Nick then – I thought there might some connection with the answer, if not the question.
Ok then, does Bilbo ever give any reason why the assumption that in the labour supply decision income effects dominate (or equal) substitution effects is any more plausible then the opposite assumption?
JKH: Yes, that’s directly related. And Bill makes the standard mistake (he is a long way from being the only one) of forgetting that for every $1 of labour sold, there’s $1 of labour bought. So if there were an exogenous change in real wages, (e.g. minimum wage laws change), there would be no aggregate direct change in income, though there would be a change in the distribution of income, and so there could be a distribution effect in either direction. (There will be indirect changes in income, of course, due to the change in quantities of employment and output due to substitution effects.)
It’s exactly the same thing, only the real wage is the relative price, not the real interest rate.
Yes Nick, you’re quite correct that Bill is using a partial equilibrium argument for a general equilibrium problem.
I was actually leading up to a slightly different point though. If you recall the original competitive labour market type models (like Lucas-Rapping) were about intertemporal labour substitution.
The upward sloping labour supply curve comes from the fact that we are tracing out labour supply vs real wage holding expectations of the future path of wages unchanged. Thus, a higher real wage is to be interpreted as higher relative to future real wages. (We could just easily talk about lowering the expected future wage.) Thus, even at the individual level there is no “income effect” because their is no implied change in lifetime income along this curve.
The upward sloping labour supply curve says nothing more then people try to maximize the present value of their lifetime utility by maximizing the present value of lifetime real labour income and the present value of lifetime utility from leisure. This maximization entails a path of labour supply that supplies more labour when real wages are relatively high and less when real wages are relatively low.
In short, this pattern of labour supply behaviour follows directly from the maximization of a lifetime utility function that derives utility from both consumption and leisure. It has nothing at all to do with assumptions regarding the relative strength of income or substitution effects because there are no income effects.
PS: The relative strength of income and substitution effects in the labour supply decision would be relevant if what we were talking about was a rise in all present and future real wages but that would constitute a shift in the entire labour supply curve, not a movement along it. In that case the relative strength of the income and substitution effects would determine in which direction the curve shifted. But this has nothing to do with the fact that the curve slopes up.
^edited to turn italics off!
Adam: “Yes Nick, you’re quite correct that Bill is using a partial equilibrium argument for a general equilibrium problem.”
That’s my point in a nutshell.
OK. I now see your point about intertemporal labour supply elasticities. Hadn’t thought of that. If current wages go up $1, and expected future wages go down $1, there’s (roughly) no income effect even at the individual level.