Recent Posts

Sunday, February 17, 2019

Comments On The Green New Deal And MMT

There is considerable interest in the Green New Deal (GND) proposal by U.S. Representative Alexandria Ocasio-Cortez, and its relationship with Modern Monetary Theory (MMT). Although my preference is to focus on my business cycle book, since I am one of the many MMT bloggers, I should at least comment on the Green New Deal. The natural question arises: what does the Green New Deal mean, and is it feasible? Based on my very limited understanding of the American legislative process, I would guess that it is way to early to say anything definitive. That said, that has not really stopped anyone else from making wild claims about the proposal...

It should be noted that the concept of a Green New Deal has attracted a good deal of attention elsewhere, such as in Europe. However, since the programme itself is uncertain, the suggested mode of implementation will vary. As a result, we can see calls for a "Green New Deal" from people who are otherwise critical of MMT. In other words, the concept is not synonymous with MMT.

A Proposal for a Proposal

At the time of writing, the proposal in the House of Representatives is just calling for the creation of a committee, who are mandated to come up with the actual set of policies. The proposal itself offers guidelines, which could be changed during the legislative process, and then the committee itself can take the ball and run it off in a different direction.

Correspondingly, any apocalyptic discussions of costs in the trillions, hyperinflation, or whatever, are just plain silly. Even if the current proposal sponsors get it passed, the end result will be the watered down result of the legislative sausage factory.

I would summarise the objective as an ambitious plan to move away from fossil fuels towards renewable energy sources. In addition to replacing the fossil fuels, it requires building an electrical grid capable of working with the intermittent nature of most renewable sources (not counting hydro-power, which is probably near maximum exploitation anyway).

I am not the person to ask whether this is technically feasible. I am in the "peak everything" camp, which is even more pessimistic than the "peak oil" crowd. I looked at the technical discussions of renewable energy years ago, and was not particularly impressed. However, I have not kept up with the state of technology nor the discussions around it, so I would suggest readers look for alternative sources for feasibility analysis.

For what it's worth, my view is that energy conservation is where progress will have to come from. The issue is that total energy demand has to be strangled. Previously, energy efficiency improvements just allowed people to increase overall consumption. To the extent that our modern standards of living are based on energy consumption (a point that was discussed in depth by Peak Oilers, when they existed), our standard of living has to be cut back. Good luck selling that as an electoral platform in a national election.


There is a wide academic Modern Monetary Theory body of research (which critics have an amazing inability to cite), including MMT academic conferences (a couple of which I have attended). Environmental issues were one subject of investigation. Since my interest was in business cycle theory, I personally never followed that literature.

From my admittedly narrow perspective, there are two key areas where the Green New Deal interacts with the MMT macro theory.
  1. The use of Job Guarantee workers to implement Green New Deal policies.
  2. How do we "pay for it?"
I will discuss these in turn.

The Job Guarantee

The Job Guarantee is a generic welfare state programme. The basic idea is that anyone who wants a job can get one via the programme, at a fixed Job Guarantee wage. This wage becomes the de facto minimum wage in the economy. (People who are unwilling to work, or unable for a number of reasons, would be expected to rely on fallback programmes.)

When the generic Job Guarantee is discussed, the most common (non-stupid) complaint of critics was a very legitimate question: what will the Job Guarantee workers do? In the worst case, paying people to stand around with shovels and telling them to look busy is not a programme that will generate goodwill among voters who are wage slaves in the private sector. The economists who have been pushing the Job Guarantee have given long lists of jobs for the pool of labour to work on; work related to the environment was always an important category.

To the extent that the Green New Deal is aimed at conservation, the Job Guarantee will be an effective policy tool. Ultimately, conservation is about changing consumption habits to be less wasteful and less energy intensive across society. The Job Guarantee workers would be educated (indoctrinated...) in new ways of thinking, and thus help spread change, on top of providing labour power for those who lack the capacity to make changes (such as insulating houses better, etc.) themselves. In wealthy areas of the country -- where the pool of Job Guarantee workers would be low -- virtue signalling would propel people to make movements towards conservation on their own dime.

The issue with a politically palatable Green New Deal is that there presumably will be particular infrastructure associated with it. The pool of labour in the Job Guarantee are the residual of private sector labour demand, and the government has no control over their geographic distribution, nor the skills available. It will be an obvious challenge to deliver particular high tech infrastructure in particular locations on a set schedule given that constraint. As a result, I think conservation efforts would be the bulk of Green New Deal work for a Job Guarantee; and other projects (such as providing labour for charities, etc,) would be the remainder of the jobs for the Job Guarantee pool.

As for the macro feasibility of a Job Guarantee, it is straightforward to see that it not a particular challenge, particularly if implemented when the unemployment rate is relatively low. There will be a one-time shock to wage scales at the bottom end of the income distribution, as minimum wage employers -- who generally are the bottom feeders of the capitalist world -- have to face real competition. Once that shock has passed, the programme is self-limiting. If the Job Guarantee spending is causing the economy to overheat, employers will be willing to hire workers out of the Job Guarantee pool at a markup over the Job Guarantee wage. This reduces the size of the Job Guarantee spending, and raises the tax take on incomes. In other words, it is self-stabilising. (This automatic stabilising disappears if we reach literal full employment, and everyone who wants a job has one in the private sector. Although this causes some hand-wringing among the people who enjoy trolling MMT, that is not a problem anyone sensible would lose sleep over.)

From a political perspective, I am unsure about the strategy of tying the Job Guarantee to the Green New Deal. My instinct is to market it as a stand alone policy to fix issues with the welfare state. Since many social conservatives objected to paying people to do nothing, it can be marketed to a relatively wide constituency. (Generating such wide political bases was a feature of Canadian Prairie Populism.)

A Rant About "Paying For"

Once we put aside the Job Guarantee, most online discussion consists of "experts" attacking various straw man versions of Modern Monetary Theory. Most of these straw men arguments are along the lines that "MMT says we just print money to pay for the Green New Deal." (Spoiler: that's not true.) One mainstream critic (who coincidentally wrote an article "explaining MMT") wrote something along these lines this week in the context of the discussion of the GND: "we need to pay for social programmes." (I am only paraphrasing since the details of the textual analysis does not really matter, nor I am going to dwell upon which luminary gave us that insight.)

If one read practically anything written by an academic MMTer, one rapidly realises that saying that we need to "pay for" programmes is a non sequitur. One branch of MMT is the study of historical monetary systems; the argument is that money was used by governments to provision themselves instead of demanding payment-in-kind. That is, rather than demand wheat from a farmer, the government demands a tax, payable in money, from the farmer. So the government will always "pay for" a programme like it will "pay for" anything else: by writing a cheque (or using an electronic transfer...). The real issue is: what are the consequences of that monetary payment?

Some people might argue that this is a semantic splitting of hairs. However, the reality is that framing the question as "how will it be paid for?" is nonsensical. (How can someone claim to be able to explain MMT if they managed to miss that?)

In order to salvage some value out of the claim, we have to assume that it was some form of appeal to the one-period governmental budget constraint. This is a mathematical expression (a constraint on a model), but can be described as an accounting identity that says that government spending is equal to the change in the money stock plus bond issuance, plus taxes. This gets bowdlerised into "government spending must be financed by taxes, bond issuance, or money printing", with the latter term being deemed equivalent to "inflation" courtesy of the Quantity Theory of Money (which is obviously rejected by the data, but whatever).

This attempt to save the logic fails. The one period budget constraint is just an accounting statement, with no causal logic behind it. The following two sets of propositions are equivalent.
  1. The government spends $N.
  2. The governmental budget constraint must hold.
  1. The government spends $N.
  2. 1-1=0.
If we want to make any statements with theoretical content, we realise that the governmental budget constraint clouds two separate issues.
  1. There is an allocation between bonds and government-issued money (the monetary base) in private sector portfolios.
  2. For a free-floating non-convertible (fiat) currency, taxes (or equivalent) are needed to maintain the purchasing power of the currency unit. Although tax policy is not the only thing that determines the price level, an absence of taxation would presumably have obvious effects. (There are not a lot of countries that have tried abolishing taxation without having alternative means of preserving their currency's value, so that is an assertion.)
If neoclassical economists actually paid attention to their own models, they would see that the money/bond allocation decision is almost of no importance. The only reason one would care if one believes in some crude Quantity Theory of Money -- a theory that is obviously rejected by the data.

We are left with the observation is that the government needs to impose taxes in order to keep a lid on the price level. However, there is no simple mathematical rule that relates fiscal variables to the price level in MMT. Although some commentators jump up and down and scream about that point, the reality is that there are no simple rules that determine the price level, other than various tautologies (the velocity of money, etc.). Neoclassical models have lots of equations that allegedly tell us about the evolution of the price level, but nobody except devoted neoclassicals use them -- because they simply do not work.

Finally, we may note that the concept of all spending being paid for by taxes is simply bad mathematics. Almost any plausible model that features positive steady state nominal GDP growth features the stock of government liabilities growing in an unbounded fashion. That stock of liabilities reflects the shortfall of government taxation versus spending. Since debt is growing in an unbounded fashion, there is literally an infinite amount of spending that has no associated tax offset.*

"Paying For" The Green New Deal

With that rant about semantics out of the way, we can turn to the more serious questions about the policy implications of the Green New Deal. There are two points of interest with respect to MMT.

  1. Should we lock interest rates at 0%?
  2. To what extend does spending need to be offset by taxes?
The first point (should we lock interest rates at 0%?) is interesting, but largely orthogonal to the question of the analysis of the GND. Even under conventional analysis, there are no strong reasons to care whether interest rates are locked at 0% for the analysis of the GND. There is room for a debate whether we should de-emphasise interest rate policy, but that is a generic question about macro policy. 

Admittedly, people like talking  about the subject of abolishing bond issuance when discussing MMT on the internet. However, just because people like talking about it means that it is important. 

The second subject is what really matters. What happens to tax rates?

From my perspective, the MMT position is straightforward. There are two parts to the position.
  1. The constraint on fiscal policy is inflation. (This is stated in practically any MMT primer.)
  2. The relationship between fiscal policy and inflation is not simple; what matters are the real constraints in the economy, and the analysis of what constraints are being hit explains the complexity of the topic. (Although some primers state this, this part of the position gets mangled in internet discussions. However, to be blunt, a lot of the poor comprehension of this is due to MMT critics being deliberately obtuse when attacking MMT.)
We can use a very straightforward example of this principle: the last round of Republican tax cuts. (Although many Republicans accuse MMTers of being left-wing radicals, the Republican party is way more MMT-influenced than the Democrats in practice.) Conventional Democrat economists screamed loudly about the risks to the "financial capacity" of the United States (a non-issue), and the risks of out-of-control inflation. The first worry contradicts the MMT position that inflation matters, not "fiscal capacity," nor the whims of "bond vigilantes." The second worry almost sounds like the MMT position. The difference was that the conventional Democrat position looked at the dollar cost of the tax cuts, and not the real resources associated with them. The Republican tax cuts did not pose inflationary risk because they ended up almost entirely in the hands of entities that would not change their spending plans: the rich and multinational corporations. The tax cut just went straight to private wealth accumulation. Portfolio rebalancing effects just meant that this implied a greater flow in Treasury securities -- "financing" the tax cut. In other words, the dollar size of the policy change told us nothing about the effect on the economy.

Back to the Green New Deal. The Green New Deal wants to make a radical change to the economy -- eliminate fossil fuel consumption, and replace it with renewable consumption. There are three legs to this change.
  1. Wiping out usage of fossil fuels.
  2. Creating new consumption patterns that use renewable energy.
  3. The one-time investment surge needed to set up the above pivot in activities.
If we ignore the third issue (investment), we are seeing a replacement of one type of private sector activity by another. Once achieved, there is no reason for the federal government to be a larger or smaller share of the national economy than before. So in the steady state, there is no particular reason for tax rates to go up or down. (If the Federal government takes a permanent role in the provision of renewable energy, that would need to be taken into account.) 

The trickier part is getting to that steady state. However, it is not a uni-directional story of massive government investment. Notice the bit about "wiping out usage of fossil fuels"? Some industrial sectors would have to be euthanised, and that is a deflationary shock. Furthermore, fossil fuels are no longer easy to extract. Considerable capital is being vapourised by firms engaging in fracking across the United States. The renewable investment will be at the cost of that existing investment into fossil fuels.

Much will depend upon the nature of the programme. If the response is to change building one type of car for another (gasoline for hybrid/electric), it is possible that the changes could be absorbed by the existing industrial capacity. However, it would be very unsurprising to run into constraints on particular inputs, such as the minerals used in battery production. Once again, the constraint is the limited production of those materials, and not the size of the national debt measured in dollars.

Furthermore, to the extent that the programme requires equipment produced elsewhere, the demand for that equipment will not pose inflationary risks for the domestic economy (beyond the trivial observation that the demand may bid up the global cost of the equipment). Such government expenditure will not be directly associated with an increase in GDP -- the import bill will subtract from domestic incomes. (One may note that domestic installation of equipment would require domestic resources, so there would be an indirect effect on domestic growth.) Certain people tend to get the vapours about the "external constraint," and intone that governments cannot undertake such policies for some reason or another. However, unless the country is pursuing a policy of autarky, such goods will have to be imported. Since that component of spending will have negligible stimulative effects on the economy, it makes no sense to raise taxes to damage private sector demand now to deal with an extremely hypothetical future risk posed by foreign holders of domestic debt. (The sensible strategy is to ignore the people who worry about external constraints, and just let the level of the currency float. If the currency weakens, the domestic economy will eventually be stimulated by a resurgence in export activity. You worry about the economy overheating when that export boom actually happens.)

(The Job Guarantee component is self-stabilising from an inflationary risk point of view, as noted earlier. If we hit capacity constraints in the labour market, that category of expenditure will drop.)

In other words, no matter what the cost of the program is, it is relatively safe bet that the required tax offset will be less the size of the investment program. You would need a detailed plan to actually attach numbers to that vague assertion. My gut feeling is that a hefty carbon tax would need to be implemented along side the investment programme, to help hasten the creative destruction of existing energy habits.

Concluding Remarks

The success or failure of a Green New Deal programme is entirely reliant upon the plausibility of a plan to replace fossil fuel consumption with renewables. What "bond vigilantes" might think about it is not something to worry about.


* Desperate defenders of orthodoxy will invoke the inter-temporal governmental budget constraint. The idea is that the discounted value of future primary surpluses are aligned against the stock of debt. The problems with this defense are manifold. The first problem is that there is no reason for the constraint to hold in the first place. Even if we put that rather large issue aside, it makes no sense to exclude interest expenses as an expense. Furthermore, the discount rate and estimated primary surpluses are essentially fictional, and there is no way of ever measuring them. Basically, saying that the inter-temporal budget constraint implies that all debt has to be paid back is as authoritative a statement as saying that debt has to be paid back because Father Christmas and the Great Pumpkin said so.

(c) Brian Romanchuk 2019


  1. I totally agree with your points about the feasibility of funding a transition. I'm very worried however about the stuff that you said was beyond the scope of your piece regarding the feasibility of decarbonising with wind+solar+tide. If the plan was to cut out CO2 using energy conservation measures (Passivhaus building standards, free public transport etc) and to build a complete fleet of nuclear power plants as France did in the 1970s and 1980s -then I'd be fully on board. As it stands, the GND looks ton have learnt tragically little from the failure of the German Energiewend where they have spent EUR 25B/year for 15years on renewables without reducing the carbon intensity of their electricity. That is as much for as long as it took France to entirely switch over to a <70gCO2/kWh electricity system using nuclear. We only get wind/sun 1/3rd of the time so a wind+solar based system won't be displacing fossil fuels. There has been pitiful progress in energy storage technology and no hope on the horizon that that will change. The UK would need 8TWh of electricity storage to enable a wind+solar+tide+storage electricity supply system. And yet we get cheering about 100MWh "futuristic" storage "solutions".

    1. Energy trends are of obvious interest to inflation-linked bond analysts, so I used to pay attention to them. And yes, I was pessimistic about the capacity for renewables/storage.

      However, we live in a society that values optimism about technological progress. There certainly has been progress for renewables since I last looked at the topic years ago. Asserting that the technology will fail would generate huge arguments. Since I want to br researching recessions and not renewable technology, I’m ducking out of that debate.

    2. You can't know how well renewables / storage or anything else would work if you don't try. I'm optimistic about it because we - humanity - haven't really even been trying. The GND is about trying. In spite of the best efforts of TPTB, Science & Technology have been advancing, albeit at a slower pace in the last 40 or 50 years. Just pick up the pace to that of the Good Old Days- by using the good old fashioned method of throwing money at it - and who knows?

    3. The Peak Oil response is that you are optimistic about technology; Peak Oilers tend to be pessimistic about technology.

      The argument is that our advances in technology are more about using *concentrated* energy better. For example, the new wind turbines could not have been built by our ancestors because they rely on modern engineering materials - with an extremely high energy density. Meawhile, total energy return on investment (EROI) is dropping in aggregate, since we are running out of the low cost petroleum sources.

      My thermodynamics course at university was utterly useless, but I believe it is safe to say that attempts to concentrate diffuse energy is directly fighting against thermodynamics laws.

      Cutting energy consumption by half is going to be a lot easier than replacing half of fossil fuel sources with renewables; it is rather telling that conservation is largely ignored in the conservation.

    4. Could not a bolt of lightning be described as a concentration of diffuse energy? Or pretty much any steam powered engine or turbine? Or even a waterwheel on a dammed river?

      You are too pessimistic Brian. There is a lot of diffuse energy out there to concentrate and exploit.

    5. Hydropower is certainly the most economical of renewable sources, but the reality is that pretty much every river in the developed world that can be dammed has been dammed. There might be areas of the world that have some untapped potential, but providing funds for hydropower was a standard shtick for development economics in the 1960s, so I am unsure how much potential is left.

      For the other sources, they take up a lot of space, and the equipment doesn't last forever. Meanwhile, they are built with modern engineering materials that are probably nearing their depletion curve peak. Technology has advanced since I looked at it, but one always needs to treat claims from technology companies with caution.

  2. That's the clearest explanation of how price level relates to various other variables I have read to date. Trying to draw unwarranted conclusions from very general statements(such as MV=PY) is far too common.

    I think one of the biggest impacts that policy can have, with regards to energy and environment issues, is that it shapes our patterns of consumption. Having jobs programs can improve the closeness of jobs to where people live, and dramatically improve the quality of those jobs. If you just want to decrease consumption, I think basic income would be intermittently beneficial, as people don't have to expend extra energy or effort to secure basic needs, even though there are significant drawbacks.


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

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

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