EconoMonitor

Ed Dolan's Econ Blog

Debt Sustainability, Growth, Interest Rates, and Inflation: Some Charts for Discussion and Some Inconvenient Truths for MMT

In a series of posts[1] [2] [3] [4] [5] over the last couple of months, fellow Economonitor blogger L. Randall Wray and I have been exploring the conditions under which the government’s debt can be said to be sustainable. Wray writes from the point of view of Modern Monetary Theory (MMT), while I adopt a more eclectic and skeptical approach.

A pivotal issue in our discussion turns out to be whether the central bank can or should hold the nominal rate of interest on government debt, R, below the rate of growth of nominal GDP, G. (We could frame the discussion in real terms instead by subtracting the rate of inflation, ΔP, from both sides; it makes no difference.) If R is held below G, then essentially any level of the government’s budget deficit is “mathematically sustainable,” a term we have been using to mean that the debt-to-GDP ratio does not grow without limit over time. On the other hand, if R exceeds G, the budget balance must show a primary surplus, on average over the business cycle, to achieve mathematical sustainability of the debt. (See the first of the posts referenced above for a detailed discussion of the conditions for mathematical sustainability.)

It seems well established that the the central bank can hold R down to any desired level, if it wants to, by buying a sufficient quantity of government securities. Barring legal restrictions like the debt ceiling, it could, if necessary, buy up all of the outstanding government debt in exchange for currency and bank reserves. Economists call this procedure “monetizing the debt.”

The “should” part of the question concerns whether the degree of monetization necessary to hold R below G would have undesirable inflationary side effects. True, when the economy is operating far below capacity and inflation is quiescent, as it has been these last few years, low interest rates and rapid money growth, backed by strong fiscal stimulus, may be just what the doctor ordered. You don’t have to subscribe to MMT to make that argument. Just read Paul Krugman. However, what happens when the economy approaches full employment and prices begin to rise? Is it still a good idea to hold R below G? That is where I become more skeptical.

So far, our discussion has proceeded largely on theoretical grounds. Now might be a useful time to supplement the theorizing with a look at the historical relationship of GDP growth to interest rates in post-World War II US experience. Doing so reveals some interesting and relevant patterns.

The following chart shows the evolution of three key variables from 1953 through 2012. The first is the inflation rate, plotted as quarterly averages of year-on-year changes in the Consumer Price Index. The second is the quarterly average interest rate on 10-year Treasury securities. (Rates on other maturities move closely together with the 10-year rate over the medium term, although shorter maturities show more quarter-to-quarter volatility.) The third variable is the growth rate of nominal GDP, plotted as quarterly observations of year-on-year changes.

A quick glance at the chart shows that the interest rate R has been sometimes above and sometimes below the growth rate G. Up to 1980, R was mostly less than G, as MMTers would like it to be. From the early 1980s through the 1990s, R mostly exceeded G. From the early 2000s to the onset of the global financial crisis, G moved back above R, but not by as much as before 1980. During the Great Recession, R has once again fallen below G.

We can see another regularity in the chart, as well. The periods when interest rates are held below the GDP growth rate tend to be periods of accelerating inflation, while those when R rises above G show decelerating inflation. The following table brings out that feature by dividing the years shown in the chart into periods broken by major turning points in the trend of inflation. The column labeled ΔΔP shows the average quarter-to-quarter acceleration (+) or deceleration (-) in the annualized rate of inflation during each period. The column labeled G-R shows the amount by which the average annual rate of NGDP growth exceeds (+) or falls short (-) of the nominal interest rate. Comparison of the two columns shows that inflation tended to accelerate during periods when G-R was positive and to slow when G-R was negative.

The apparent regularities shown in the chart and table have a direct bearing on the debate over sustainability of fiscal policy. MMT proponents argue that the interest rate on government debt is a policy variable that the central bank of a country with a sovereign, floating-rate currency can set wherever it wants. As long as R is held below G, they say, the debt can never become mathematically unsustainable. By holding R permanently below G, the government can focus fiscal policy on achieving full employment without worrying about the budget deficit or the debt ratio. On the other hand, if a policy of holding R below G produces accelerating inflation as a side effect, it starts to look less attractive, at least to observers outside the MMT camp.

Let’s now add another variable, unemployment, to the discussion. The next pair of charts, like the familiar Phillips curve, plot unemployment on the horizontal axis and inflation on the vertical axis. However, the patterns in these charts look nothing like the classic Phillips curve, which posited a smooth menu along which policymakers could choose lower unemployment at the expense of a bit of inflation, or vice-versa.

Instead, the left-hand chart, which covers 1961-1982, shows an unstable stop-go cycle with a distinct inflationary bias. Each repetition of the cycle reaches a higher inflation maximum and turns up again from a higher inflation minimum. The minimum and maximum values of unemployment also increase with each cycle. The stop-go pattern coincides with the period during which interest rates on the federal debt averaged less than the growth rate of nominal GDP.

After 1982, the pattern changes sharply, as shown in the right-hand chart. There is a transition period in the mid-to late 1980s, after which, from the early 1990s up to the start of the global crisis, the economy exhibits remarkable stability. These were the years we call the Great Moderation. The transition and the early part of the Great Moderation coincided with a level of R significantly above G. After 2002, G rose above R once again. Most readers are probably familiar with the argument that the Fed’s attempt in the mid-2000s to prolong the Great Moderation by holding interest rates low contributed to the housing bubble and set the stage for the crisis.

Of course, these are only charts. They show associations, not causation. Even so, the charts by themselves pose an inconvenient truth for proponents of MMT, in much the same way that charts showing an association between atmospheric CO2 and rising global temperatures pose an inconvenient truth for climate skeptics. In both cases, the hypotheses of causation, which leap out from charts even for naïve observers, are backed by bodies of theory that many experts in the fields find plausible. The associations, backed by plausible theory, are not irrefutable, but they do shift the burden of proof.

The question stands then: Is it not just possible but also desirable for the central bank to hold nominal interest rates on government debt below the rate of nominal GDP growth, and to do so not just as a temporary measure during a deep slump, but as a long-term policy? Do past episodes when low interest rates were associated with accelerating inflation and stop-go instability provide grounds for caution, or does each such episode have some benign, extrinsic explanation? I look forward to learning more about how MMT proponents interpret the features of the charts presented here.

55 Responses to “Debt Sustainability, Growth, Interest Rates, and Inflation: Some Charts for Discussion and Some Inconvenient Truths for MMT”

ThomasGrennesJanuary 28th, 2013 at 11:19 am

Governments can target interest rates or inflation rates, but not both. By targeting interest rates, governments lose control over the the quantity of bonds they must buy and the quantity of new money they must exchange for bonds.Given the demand for money, losing control over the supply of money implies losing control over inflation. Inflation is a tax on all holders of money, and history is filled with attempts by citizens to protect themselves against governments that imposed inflation taxes. Requiring that money be convertible into gold at a fixed price was an attempt to constrain governments under commodity money. Under fiat money, imposing an inflation target places a similar restraint on the government. Thus, the desirability of fixing interest rates depends on the optimal inflation rate for a country. The recent hyperinflation in Zimbabwe is an extreme example of a government attempting to obtain all of its revenue by taxing domestic money holders.

kakatoaJanuary 28th, 2013 at 11:46 am

I recall that how inflation is calculated changed during the Clinton administration. By chance in graphs above has some normalization of the data been done to account for the different methodologies used over time to calculate the inflation rate?

EdDolanJanuary 28th, 2013 at 12:20 pm

There have been many changes over the years in the way inflation is calculated. Yes, there are some reasons to think that in decades past, the CPI overstated inflation rates because of inadequate quality adjustments, the substitution bias that arises from using base-year quantity weights, and the way housing expenses were calculated, among other things. Such biases are presumably mitigated by the new methodology. However, for two reasons, it is unlikely that the changes would affect the patterns shown.

First, the changes, by and large, make a difference of only a fraction of a percent in yearly inflation rates.

Second, although the changes might fractionally affect the annual rates of inflation shown, it is unlikely that they would affect the medium term trends of acceleration or deceleration–the rate at which the rate of inflation changes–that are the central focus of this post.

For a cross-check, you could chart the variables using other inflation measures, for example, producer price indexes, the GDP deflator, or the index of personal consumption expenditures in GDP. Those use different methodologies, and although perhaps affected by biases of their own, they would be different biases. I'll try it myself if I can find time. Might make material for another post.

You could even try using the inflation measure from Shadowstats for the post-1990s period. That series, which has very low credibility among economists, has risen faster than the official CPI. Using it would no doubt reinforce the notion that low interest rates cause rapid inflation.

Redhwan nomanJanuary 28th, 2013 at 1:58 pm

On the other hand, if R exceeds G, the budget balance must show a primary surplus, on average over the business cycle, to achieve mathematical sustainability of the ???? how this come when the government pay more on interests burden or cost of debt than GDP growth ????
True, when the economy is operating far below capacity and inflation is quiescent,….??!! Paul Samuelson mentioned in his book that inflation doesn't occur when below capacity it is happened in full capacity employed

EdDolanJanuary 28th, 2013 at 2:27 pm

With regard to the R>G question and primary balance, see reference [1] at the beginning of the post, including the appended slideshow

With regard to "below capacity and inflation is quiescent", my meaning was similar to what Samuelson's textbook says, namely, that we don't have to worry too much about inflation during periods when the economy is below capacity.

David WishartJanuary 28th, 2013 at 3:05 pm

Another interpretation of the chart is that when the growth rate exceeded the interest rate in the early period, the U.S. experienced higher levels of real GDP growth, while in the latter period, when interest was higher than the nominal growth rate, real GDP growth was smaller. The OPEC oil shortages of the 1970s created an inflation spike that had nothing to do with domestic monetary or fiscal policy, so I'd be uneasy about including these years as an indication of interest rates on inflation.

David WishartJanuary 28th, 2013 at 3:05 pm

Expansions in private sector credit, new money created through private lending (debt) has an interest expense attached. To cover the interest expense, you increase prices where possible. So the debt supercycle of the last 30 years as interest rates have fallen has imbedded a huge amount of interest expense in every product and service's price. It's possible then that an increase in interest rates against an economy with too much private debt will result in an inflationary spike. I'm not convinced that low interest rates cause inflation, but rather that there is too much private bank credit money relative to non-debt backed government money. The Vietnam War appears to have led to some acceleration in inflation, but otherwise, the early period had very low inflation as the US invested in productive capacity while the overall growth of inflation has remained persistently around 3 percent as private debts have swelled.

kakatoaJanuary 28th, 2013 at 3:31 pm

Thanks for the clarification Ed.

It's been a few, ok make that a lot, of years since I spent any time in the details of econometric models. I concur with your point two that it really wouldn't make any difference to the trends which way the official rate was calculated (the slopes in the graphs as large enough to swamp out small errors in the inputs).

EdDolanJanuary 28th, 2013 at 4:09 pm

Yes, supply shocks can matter. However, the conventional view is that the amount of inflation you get from a real supply shock depends both on the nature of the shock itself and on the degree to which policy accommodates the shock. It is not quite right, then, to speak of "an inflation spike that had nothing to do with domestic monetary or fiscal policy." There could only be a shock that had "nothing to do" with domestic policy if neither monetary nor fiscal policy had any effect at all on nominal aggregate demand.

EdDolanJanuary 28th, 2013 at 6:09 pm

The Chinese economy is certainly doing well in many respects, but I am not sure that huge investments in projects that return less than the cost of capital is one of its strengths. True, the central government can paper over some of those bad debts, especially since it has very little conventional debt to begin with. Meanwhile, the Chinese financial system does not do a good job of prioritizing financial flows for the most productive investments, and Chinese consumers have a very low standard of living, considering the size of the economy.

Bryan WillmanJanuary 28th, 2013 at 6:48 pm

Let's try an extreme thought experiment, and then wonder what if anything it says about much more moderated reality.

Imagine that the central bank monetizes debt without bound, and that government uses this to engage in extreme deficit spending – let's suppose govt runs a primary deficit of 100% of GDP.

Imagine further that none of that deficit is hidden away in trust funds or other such gimicks, and that virtual none of it is invested in government debt nor buried in tin cans – it all enters the real economy.

In this thought experiment R is low because the central bank is printing money at light speed. But the supply of money bidding for the real output of the economy is 100% of GDP – real growth. If real growth is, say, 25%, then the effect would appear to be 75% inflation per year, no?

It would naively seem that ANY time the supply of money in circulation which is in effect bidding for the output of the economy is growing faster than the real economy, this will cause inflation. If this is not so, please explain why not?

SchofieldJanuary 28th, 2013 at 6:56 pm

This obsession with debt in the West is an illness. For thirty odd years now the Chinese central bank has been making good the non-performing loan portfolio of state owned local banks. For thirty odd years Chinese GDP has been averaging up to 10% annual growth until the Western contrived Great Recession arrived leaving the Chinese with a current GDP growth rate of 7.8% and growing. Perhaps if the West made the effort to remember what money has been primarily created to do and the word "instruction" wouldn't go amiss then the possibility of recovery from "debt illness" might be possible.

JohnJanuary 28th, 2013 at 7:08 pm

The above piece does raise some interesting questions, but in my view, they fail to undermine MMT in any way. The clue is in the article itself. The graphs show an 'association, not causation'. The economy is a complex interaction of many factors; it would be nice if these were themselves independent, but most, if not all major factors, are inter-related. I suggest that unscrambling all the factors to show causation requires a far more detailed, and probably mathematical analysis than is traditionally attempted in a blog. (I would love to see it though). I just wonder about the fact that the policies used to control the economy throughout this period have been of the neo liberal tendency. The control measures used, both monetary and fiscal, have not been appropriate for the way that the economy actually works. What effect does this have on the output data from the economy? Could the associations be caused by the chosen levers that are pulled by ill-informed politicians and economists producing these results? I also wonder about all the other factors / levers that have been used over this long period, what about changes in taxes, both local and federal? So yes, there may be an association, but the data presented here is a long way from even suggesting that MMT needs to defend itself. It does indicate a likely fruitful area of future research in which MMT proponents need to be fully and actively involved with.

EdDolanJanuary 28th, 2013 at 7:47 pm

I think that to make the numbers in this kind of example work, we would have to be a little more precise about the difference between the quantity of money (the stock of outstanding currency and bank deposits) and the rate of spending (the flow of government purchases of goods and services, consumption, investment, and exports). Even so, I think your intuition is right that if total spending (economists would call it nominal GDP) grows faster than total output (economists would call it real GDP), then you are going to get inflation.

EdDolanJanuary 28th, 2013 at 7:51 pm

What I am hoping is not so much to inspire additional research my MMT proponents, but to get them to lay out their thinking in a way that is easily understood by non-MMT economists. It is not as easy as it should be to set the two bodies of theory side by side, and say "Oh, THAT's obviously the key assumption or parameter that leads to different policy conclusions.

Bryan WillmanJanuary 28th, 2013 at 8:38 pm

Yes.

I conjecture that a very high percentage of government spending, whether purchasing of output that govt. then distributes, or transfers to citizens to do with as they wish, ends up as GDP rather than stocks of money. (By illustration – I assert that most social security recipients spend most or all of their SS payments.) I guess one could try to measure that, and said measurement would inform the NGDP/RGDP effects pretty directly.

This would be a case where "stimulus" might "look good" (it increased NGDP) but might not actually be good (it might all get eaten as inflation) – depending on the slack in the economy when the stimulus was attempted.

So I think I agree with your post – keeping R < G does not assure a sustainable result.

Bryan WillmanJanuary 28th, 2013 at 8:42 pm

To understand this don't we have to sort out all the covariables?

Over such a period, has the perceived value of leisure over work changed? Have markets improved or become more distorted in macro-economically important ways? Has the effective level of financial repression changed? Has the ratio of GDP to total real wealth in society changed?

JohnEJanuary 28th, 2013 at 9:59 pm

My reading of MMT is that it is more a description of the implications of having a 1) fiat
currency 2)international floating currency exchange rates and 3) all sovereign
debts denominated in the domestic fiat currency. The implications seem to be that
1)government is self-financing,2)taxes serve to enforce the government's monopoly on said fiat currency, 3)that the bond market is an alternate channel to introduce currency into circulation by paying interest on deposits and 4)there is no solvency issue.
I would agree that MMT seems to say that the natural rate of interest is zero.
The implication is that regulation of the economy should be by raising or lowering
fiscal deficits to control inflation, which should be done by Congress; as opposed to
interest rate changes by the Federal Reserve Bank. MMT does worry about inflation;
MMT may be overly optimistic regarding the understanding/ability of Congress to
regulate the deficit to regulate inflation. If Congress screws up, you will be right; but
that doesn't mean that MMT is wrong. It will just prove how stupid 100 Senators
and 435 Reps can be. Given the current deficit obsession, I doubt inflation will be the
problem. As Warren Mosler says, because we think we are turning into Greece
we will instead become like Japan with chronic subpar growth.

EEBJanuary 29th, 2013 at 4:41 am

NO!!! The Zimbabwe hyperinflation was caused by the fact that that country needed hard currency to pay for its imports, but could not run sufficiently high trade surpluses to earn the hard currency needed for that purpose. Since Zimbabwe is not exactly an export powerhouse, the only way it could obtain hard currency was in the forex markets, so it dumped ever larger quantities of its soft money in the forex markets to buy hard currency, mostly U.S. dollars. This then drove the Zimbabwe currency down and set off the hyperinflation, beginning with import price inflation. But now that Zimbabwe has dollarized, it won't be long before it experiences Austerity, Deflation, and Depression when it finds that its debts will be denominated in a hard currency that they still will not be able to amass in sufficient quantities to pay for their imports.

EEBJanuary 29th, 2013 at 4:54 am

Well, that would depend on the relationship between actual and potential GDP in the initial condition. If this regime were instituted during the bottom of a massive depression, then it is most probable that there would be 100% real GDP growth to match; of course, as they say, 100% of nothing is- well, nothing! On the other hand, if the economy were at full employment and actual GDP and potential GDP were equal, then this thought experiment would eventually produce hyperinflation unless(perhaps confiscatory) taxes were raised, or the economy's potential were increased. Intermediate initial conditions might yield a variety of outcomes, depending on many other variables, including labor plroductivity and the nature of industry cost curves(e.g., increasing, decreasing, or constant returns to scale).

JohnJanuary 29th, 2013 at 8:17 am

It would be good to see this happen, I agree. One of the problems with economics as a subject at almost any level is that everything seems to be up for debate. There are a few 'facts', but even here the debate then starts on their relevance/importance.

The audience for any debate needs to be willing to listen in the first place. In my view, very few people are willing to even consider having their preconceived notions altered or overturned.

David WishartJanuary 29th, 2013 at 10:08 am

Good point, there are definitely some dynamic interrelationships between a supply shock and the reaction to that shock by policy makers. Perhaps policy is mostly impotent in this scenariom, aside from massive investment in alternative supply sources. Would inflation have settled down anyway after the market started producing more alternatives to OPEC produced oil. Was the reduction in inflation and Volcker's rate hikes merely coincidental, and have occurred anyway without the monetary intervention?

David WishartJanuary 29th, 2013 at 10:16 am

I would be interested in seeing the employment growth rate in the chart as well to see how it fits in to the story. The Philips curves show falling inflation causing unemployment to rise after 1970s, but rising inflation not to cause unemployment to fall, the stagflation scenario. But since then it's hard to glean any kind of relationship at all, so it would be interesting to see how employment growth looks across the longer time series. At this point the discussion will likely start to drift towards the relative merits of a higher inflation target to achieve higher employment and it starts to turn to some degree towards a value judgment.

EdDolanJanuary 29th, 2013 at 10:20 am

I don't see that your MMT analysis (closely related to Bill Mitchell's analysis, I guess) of Zimbabwe's inflation differs much from Tom's analysis. Both of you are saying that the government created an excess supply of Zimbabwe dollars, and that caused inflation. Am I wrong?

RuiJanuary 29th, 2013 at 11:00 am

Agree with Ed about this. Someone has to pay the bill, if they bail the banks and local goverments this doesn't mean it doesn't come with a cost. I am not sure what your point is. Seems like you are saying that goverment and state owned banks in China (or any other country) can inflate economy at zero cost, which I think is not true.
China has experienced quick development and fast growth in last 30 years, usually with huge trade balance surplus. This means they have been piling up tons of foreign reserve (dollars), which normally will mean they are accumulating wealth.
But since the FED insist in they idea of QE (which takes value from dollar making deficit and debt cheaper in real terms) they will find that these reserves are not worth as much as they use to be. Also someday they will have to pay the bill of all malinvestment, so I think the standard of living in the next years will not rise up as expected, at least for the majority of people. Seems China is going the same road as Japan did before, both seem unstoppable at their moment.

EdDolanJanuary 29th, 2013 at 11:14 am

You comment, "MMT does worry about inflation; MMT may be overly optimistic regarding the understanding/ability of Congress to regulate the deficit to regulate inflation."

I am concerned about that as well. In my original post that started this discussion ( http://www.economonitor.com/dolanecon/2012/11/28/… ) I included "functional sustainability" as the third, and most important, meaning of sustainability. My definition was this:

"If a country has a set of rules and decision-making procedures that adjust fiscal parameters over time to serve some rational public purpose, we can say that its fiscal policy is functionally sustainable."

My discussion with Wray has not even reached that part of the problem yet.

Here is the real nightmare. What if it turns out to be true that MMT is right in saying that the central bank is powerless to affect aggregate demand one way or another (no control over reserves, no interest sensitivity of investment, completely elastic demand for money balances, etc.). At the same time, suppose it turns out that at least for the US, fiscal policy is functionally unsustainable. Then there is just no hope. It is Greece, Japan, or Zimbabwe ahead, we just don't know which yet.

"Some say the world will end in fire, some say in ice . . . "

JohnEJanuary 29th, 2013 at 11:57 am

Definitely not Greece. They don't control their currency.
Hard to imagine Zimbabwe unless all the deficit talk in DC is bluster.
Japan likely: High standard of living for those with jobs and malaise for those without.
Our one advantage is growing population. Demographics matter.

(Enjoyed your course at Dartmouth in microeconomics for which you gave me
a faculty citation, I think I wrote about contraception lowering the perceived costs
of intercourse and possible unintended consequences.)
D "78

john1025January 29th, 2013 at 12:09 pm

All hyperinflations in the past 200+ years have begun collapses in production, rampant government corruption, loss of a war, regime change or collapse, pegged currency rates or foreign denominated debts. Rampant money creation comes at the tail end of one or more of these events, not at the beginning. What happened in Zimbabwe arose from land reallocations which sliced up their largest export and domestic form of productivity into the hands of the incompetent. Internal food production collapsed and the government was forced to rely on food imports and IMF aid. Unemployment increased, civil unrest increased and the government lost control of its internal finances and the currency ultimately collapsed as the people voted no confidence in the currency. All of this led to hyperinflaton.

We are far from an inflationary event in the United States as there is a huge economic output and capacity, high unemployment and underemployment, and stagnant wages. When the government spends it is not adding net financial assets. It spend by marking up a recipient’s bank’s reserve account and the bank marks up the recipient’s bank deposit account. The bank’s balance sheet has changed in composition –more reserves (assets), more deposits (liabilities) – but no net change in financial assets.

ThomasGrennesJanuary 29th, 2013 at 7:02 pm

China is a bad example for this point. Yes, the central governmet has supported regional governments and this support can be interpreted as implicit debt. However, the Chinese government has enormous gross assets that swamp these implicit debts. China has the largest stock of monetary reserves in the world, and the largest component is denominated in U.S. dollars. China is a large net creditor.

ThomasGrennesJanuary 29th, 2013 at 9:11 pm

China is a particularly bad example for this point. True, the federal government has acquired some implicit debt by assisting local banks and governments. But the Chinese
government's gross assets completely swamp these liabilities. China has the world's
largest stock of monetary reserves, and they are the single largest holder of U.S. government bonds. China is not a net debtor, and its rapid growth cannot be attributed to debt.

ThomasGrennesJanuary 29th, 2013 at 9:17 pm

If Zimbabwe's hyper-inflation was attributable to real economic shocks (smaller crops, etc)
rather than excess creation of Zimbabwean money, why did the inflation stop when the old currency was abandoned?

OliverJanuary 30th, 2013 at 8:10 am

So if I read correctly, you state that the period in which R<G, as you both say would be necessary for the debt/gdp ratio to be sustainable while running gvt. deficits without offsetting external surpluses, showed cyclical bouts of inflation and unemplyoment, as well as a secular upward trend of both which were only 'cured' once R>G was put in place beginning of the '80s?

The standard Post Keynesian explanation for the '70s stagflation is a mixture of supply shock (oil shock) and institutionalised wage arrangements (unionisation, automatic inflation adjusment etc.), Here is a good summary of this point of view:
http://socialdemocracy21stcentury.blogspot.ch/201…

I don't see how your post puts a dent in this line of explanation. To me, the first graph shows the CB reaction function and says little about causation.

OliverJanuary 30th, 2013 at 8:13 am

But to put things into relation, I don't think PKers / MMTers claim that interest rates have no influence on macro measures at all, just that there are better methods of dealing with macro problems than tweaking interest rates. They are die hard fiscalites, as I'm sure you've noticed.

This has a number of reasons:

- Business investment is not interest rate sensitive. In fact, businesses do not borrow to invest in production much at all any more. So, interest rates do not work to regulate investment and consumption.

- The channel by which the CB target rate does work is via mortgages. On the one hand by discouraging new borrowing, on the other by punishing existing mortage holders. Bank lending is pro cyclical, or to put it differently, mortgages are the cycle.

- The amount by which rates must be raised to break the cycle tends to be so great that it necessarily causes an 'artificial' recession. It's a very broad brush to be painting with and effectivley, via the mortgage channel, means that housholds are put first in line to make cyclical adjustments.

OliverJanuary 30th, 2013 at 8:14 am

- The whole framework of your argument rests on an 'all else equal' assumption in the sense that other channels and other levers for regulation are implicitly denied. Part of MMT rhetoric tactics is to break that dominance of monetary over all other levers and force a discussion of all available possibilities.

- Interest has an effect on borrowing but also constitutes a net income channel to non government via interest paid on government (and/or CB) liabilities. This income and its distributional implications need to be accounted for when comparing it to other possible sources of income. Roughly speaking, interest on gvt. liabilities is income for natural savers and arbitrageurs (pension funds, sovereign wealth funds, financial institutions) as opposed to income of wage earners that feeds more directly into consumption. I.e. it's a worker vs. capitalists kind of thing.

and probably a number of other arguments I haven't thought of.

P.S. R<G is not a sufficient condition for stable debt / GDP ratios, nor does R>G automatically lead to a debt crisis: http://www.concertedaction.com/2013/01/17/claims-…

EdDolanJanuary 30th, 2013 at 3:31 pm

Thanks for your three comments. Let me put all my rejoinders in one place.

1. "The standard Post Keynesian explanation for the '70s stagflation is a mixture of supply shock (oil shock) and institutionalised wage arrangements (unionisation, automatic inflation adjusment etc.) . . . To me, the first graph shows the CB reaction function and says little about causation."

Of course I agree that inflation in any given period depends both on external shocks and on how the central bank reacts to them. The issue here is that the only time we saw R<G in the last 60 years was at a time when the combined effect of shocks and CB reaction were causing accelerating inflation. That still leaves the question hanging: Is it just coincidence that R<G in this period, or does it have something to do with the way the CB reacted? For example, what if it had, from the early 70s, reacted to target inflation at 2% in the face of the oil shocks? I still need an answer to that question.

2. "But to put things into relation, I don't think PKers / MMTers claim that interest rates have no influence on macro measures at all, just that there are better methods of dealing with macro problems than tweaking interest rates. "

Well, I'm happy to hear that. Some MMTers talk a kind of shorthand that would lead to think they see no effect of interest rates on investment at all, even housing.

3. "The whole framework of your argument rests on an 'all else equal' assumption in the sense that other channels and other levers for regulation are implicitly denied."

I wasn't sure what you were talking about until I read the linked item at Concerted Action. Here is my response:

First, I think both you and the CA piece are mixed up. It looks to me that you are both arguing that R<G is not a NECESSARY condition for a stable debt ratio. I agree with that. You can also get a stable debt ratio when R>G and you have an appropriate structural primary surplus.

Second, I fully agree about the importance of automatic stabilizers, and I am aware that growth of GDP causes a decrease in the deficit. To adjust for that effect, I state the conditions for sustainability in term of the structural primary budget balance, which by definition fully takes automatic fiscal stabilizers into account. It remains true that with R>G and with a positive starting debt, you need a structural primary surplus to avoid mathematically unsustainable debt ratio.

OliverJanuary 31st, 2013 at 5:03 am

That still leaves the question hanging: Is it just coincidence that R<G in this period, or does it have something to do with the way the CB reacted? For example, what if it had, from the early 70s, reacted to target inflation at 2% in the face of the oil shocks? I still need an answer to that question. </i>

To me, the correct way to phrase both questions would have to be: Assuming that it was R>G that caused the accelerating inflation in this period, is there a conceivable policy stance that could have prevented this outcome with the CB running the same reaction function (and the same structural primary deficit)?

OliverJanuary 31st, 2013 at 5:06 am

I think the MMT answer would be 'yes'. Or at least I think they would be confident in stating that there are a lot of other things that could be done to avoid the point at which it might be sensible for the CB to step in with a rate hike (the hand brake, so to speak). The main policy lever being the way in which wages and pensions are adjusted to inflation – that's what causes the spiralling. But apart from that, I think the MMT position is that cost push shocks cause prices to rise (duh), period. There's absolutely nothing a CB can do about it. And the subsequent rise in wages is a natural reaction with purely distributional effects, at least to the extent that it doesn't cause a feedback loop. In the case of the '70s, it was the feedback that needed to be addressed, not the initial rise in prices.

So, to answer your first question: no, it needn't be coincidence that R<G in this period, but MMTers claim that their analysis offers alternative remedies to those of Paul Volcker and Maggie Thatcher. That's what I meant with the 'all else equal assumption'. I could rephrase that as TIAA (there is an alternative).

OliverJanuary 31st, 2013 at 6:15 am

forget my reply above, I'll start again:

1(3)

That still leaves the question hanging: Is it just coincidence that R<G in this period, or does it have something to do with the way the CB reacted? For example, what if it had, from the early 70s, reacted to target inflation at 2% in the face of the oil shocks? I still need an answer to that question. </i>

To me, the correct way to phrase both questions would have to be: Assuming that it was R>G that caused the accelerating inflation in this period, is there a conceivable policy stance that could have prevented this outcome with the CB running the same reaction function (and the same structural primary deficit)?

OliverJanuary 31st, 2013 at 12:43 pm

And to come back to your repeated concern about the primary structural balance, I tend to think of it this way: there three sources of liquidity for an economy: 1. net exports 2. a budget deficit 3. credit. The mainstream aproach seems to be: aim for a large 1. (preferably for all nations ;-) ), keep 2. at a primary balance over the cycle and keep credit bubbles a blowing and tweak or prick them with interest rates so that 3. is as large as it needs to be for whichever measure seems improten (inflation, unemployment). The MMT approach could be paraphrased as: 1. is wherever it is (ideally in deficit for wealthier nations and in surplus for developing nations), 2. should be in primary deficit (for nations running an external deficit) so as to enable full employment and 3. is to finance productive investment only but should be kept at bay not primarily with interest rates but through regulatory reform. Note also that, the larger the deficit of 2., the less need there is for 3. overall. Thus, cyclical and anti-cyclical sources of liquidity can be brought into a better / more stable balance. A balanced budget over the cycle also means you need an ever growing amount of credit or exports as a percentage of GDP!

OliverJanuary 31st, 2013 at 12:59 pm

Sorry for spamming you, Ed. Something was blocked from the office, so I tried again from home with the above result. Feel free to delete and rearrange.

EdDolanJanuary 31st, 2013 at 2:04 pm

Here is something I am having a hard time understanding about MMT. Maybe you can clarify.

Sometimes MMTers say things that suggest that monetary policy has little impact on aggregate demand, prices, output, etc., or to put it in economic jargon, that the transmission mechanism for monetary policy is very weak. For example, your statement " think the MMT position is that cost push shocks cause prices to rise (duh), period. There's absolutely nothing a CB can do about it."

Other times MMTers write things that suggest that monetary policy mistakes can be very damaging. For example, you statement "there are a lot of other things that could be done to avoid the point at which it might be sensible for the CB to step in with a rate hike (the hand brake, so to speak)," or your implied comment that Volcker's policies were harmful to the economy.

Here is the question then: If the transmission mechanism is weak and the CB can't do anything to prevent supply shocks from causing inflation, then how can the transmission mechanism also be a harmful "handbrake" to be applied or misapplied by someone like Volcker?

What am I missing?

You write "there are a lot of other things that could be done to avoid the point at which it might be sensible for the CB to step in with a rate hike (the hand brake, so to speak) . . The main policy lever being the way in which wages and pensions are adjusted to inflation – that's what causes the spiralling. But apart from that, I think the MMT position is that cost push shocks cause prices to rise (duh), period. There's absolutely nothing a CB can do about it.

OliverFebruary 1st, 2013 at 7:24 am

OK, let me try. I've think there are 2 important distinctions to be made: that between cost push and demand pull inflation and that between consumer price inflation (i.e. inflation) and asset price inflation (i.e. bubbles).

An external cost push, say through an Opec or Saudi price hike, is a unilateral decision by a cartel or swing producer and can't be prevented by a national central bank. Do you agree? That's what I meant with 'There's absolutely nothing a CB can do about it'. And since most oil is paid for in $, not even an exchange rate intervention can mitigate the effects. Further, since oil is an input in a large amopunt of goods as well as being a fuel, price changes will ripple through most all consumer prices. Effectively an oil price shock is a deterioration of the terms of trade. There can then be third and fourth rate effects on wages that can and should be addressed locally…

OliverFebruary 1st, 2013 at 7:25 am

…Domestic credit bubbles are a different story all together. Personally, I would say the transmission mechanism for monetary policy to domestic asset prices, especially real estate, is fairly strong. Effects on consumer prices are second order effects because the marginal $ pumped into an asset bubble gets stuck in the assets themselves. And it still begs the question why one would willingly design the system such that it promotes the creation of bubbles instead of preventing them.

OliverFebruary 1st, 2013 at 7:37 am

That doesn't answer the question about Volcker though, because he wasn't pricking a credit bubble. The standard account is that he broke through the inflation spiral by inducing a heavy recession in which debt deflation and unemployment brought domestic prices down. I agree that my accounts above do not offer an alternative. But, instead of asking 'was it harmful or good for the economy?', I would be interested to hear your account of who, exactly, you believed profited from the intervention and who didn't, whether it had a short term or long term effect and, if your answer is 'long term', then in what sense it paved the way for what later became known as 'the great moderation'? http://www.interfluidity.com/posts/1256656346.sht…

BWildsMarch 12th, 2013 at 9:05 am

The 1980 recession was self induced to stop inflation by raising interest
rates. As soon as the "fever" ended and rates began to drop growth resumed.
That is not so in this case. More troubling is this each of the recent post
recovery periods following a recession is showing a slower bounce back.
The worst part of the problem today is how ugly the numbers have become,
I have posted several articles on my blog concerning this subject. Let us
hope Washington does not use another cheap trick to sidestep its fiscal
responsibilities;
http://brucewilds.blogspot.com/2013/01/ugly-math-…

HepionMarch 21st, 2013 at 12:27 pm

Thomas says that government can "loose control over money supply by buying bonds." This is wrong because governments do not control money supply. Banks do not lend out reserves.

Here is Peter Stella:

From journalists, former colleagues, professors at Harvard, Yale and Columbia. I have been reading similar ideas/commentary for almost 5 years. That is, somehow bank reserves at the central bank ought to be “lent out”, i.e. should exit the “vault” of the BOE, Fed or ECB and begin circulating in the economy. The obverse of this is that an increase in excess reserves at the central bank reflects commercial banks “hoarding” liquidity rather than lending it “out”.

Naturally, this stunningly incorrect conceptualization of the lending process and how it interacts with bank reserves leads people to think how to entice banks to “get this money out the door” including to thoughts of “negative” deposit rates as an incentive.

My frustration lies in my inability to explain to “sophisticated” people why in a modern monetary system–fiat money, floating exchange rate world–there is absolutely no correlation between bank reserves and lending. And, more fundamentally, that banks do not lend “reserves”.
http://ftalphaville.ft.com/2012/07/03/1067591/the…

HepionMarch 21st, 2013 at 12:52 pm

Hyperinflation in Zimbabwe was caused by foreign currency purchases. After agricultural production dropped and they had starving people they just tried to purchase FX to import food.

Warren Mosler calls foreign currency purchases "off-balance sheet deficit spending", because it does not show up budget deficits even though it is functionally the same thing.

So budged deficits were hyper-large, in case of a weimar that tired to do same thing up to 50% of GDP.

Budget deficits, if they are too large, can cause inflation, says MMT.

vimothyMarch 24th, 2013 at 4:05 pm

>>It seems well established that the the central bank can hold R down to any desired level, if it wants to, by buying a sufficient quantity of government securities.

How does this enable the government to set R at any level? At some point the central holds all of the treasury bonds and the government borrows by issuing money. But then R is determined by the rate of inflation or deflation.

SchofieldMarch 26th, 2013 at 3:02 pm

Ed you are having a pot calls the kettle black moment! Here we are in the West still suffering from a Great Recession caused by a bunch of greedy bankers engaging in massive fraud necessitating the spending of trillions of dollars to avoid a Credit Melt-Down and with no bankers in jail and you talk about Chinese malinvestment! The reality is that the creation of money to use as a technology will always be vulnerable to malinvestment because money is simply a way of storing the ability to instruct labour to interact with resources to produce goods and services and we can make bad judgement in regard to what we instruct whether it be for public or private use. Add to that the ease with which this "monetized instruction" can be created by government or private bankers and there's no utopia to be had. Here's a recognition of that in China:-
http://www.macrobusiness.com.au/2013/03/china-alr…

Nathan TankusMarch 30th, 2013 at 8:46 pm

this entire post is such a cop out. you seem to be too afraid to actually state your casual mechanism for fear that others will refute it. instead you want to get people to attack a correlation. you should either post a causal mechanism you think contradicts what neo-chartalists are saying, or post it as a question, not as a "criticism". a correlation isn't a criticism, it's laziness.

EdDolanApril 2nd, 2013 at 10:08 am

I am sorry that you did not like my rhetorical device of posing questions to be answered. However, if you read the post carefully, along with the comments, you will find that the causal mechanism that I am asking MMTers to address is fairly clear. The mechanism posits that the price investors are willing to pay for government bonds, and hence their nominal yields, depend on the expected real rate over their period to maturity and the expected rate of inflation over the same period. In periods of unexpectedly accelerating inflation, nominal yields will lag inflation. That sets up a potential positive feedback loop: accommodative monetary policy stimulates aggregate demand, rising aggregate demand causes inflation to accelerate, accelerating inflation causes yields to lag behind current inflation. My question to MMTers is to explain more clearly just which link in the causal mechanism they disagree with.

senexxApril 6th, 2013 at 1:15 am

The argument here for functional sustainability would be whether those rules and decision making procedures truly do serve public purpose, no matter what the rules are, that will be debatable.

senexxApril 6th, 2013 at 1:22 am

I think that helps to clarify your position. If the argument was just that r>g, I would have noted that largely coincides with the Volcker period which Randy has already addressed.

senexxApril 6th, 2013 at 2:01 am

There's every chance I've misunderstood but this post here seems to be asking the same question from the opposite point of view.

"Those who think low interest rates are inevitably inflationary need to explain the monetary transmissions mechanism they have in mind."

Most Read | Featured | Popular

Blogger Spotlight

Dan Steinbock

Dr Dan Steinbock is a recognized expert of the multipolar world. He focuses on international business, international relations, investment and risk among the major advanced economies (G7) and large emerging economies (BRICS and beyond). In addition to his advisory activities (www.differencegroup.net), he is affiliated with major US universities as well as international think-tanks, such as India China and America Institute (USA), Shanghai Institutes for International Studies (China) and EU Center (Singapore).

Economics Blog Aggregator

Our favorite economics blogs aggregated.