Recent Posts

Wednesday, May 6, 2020

Quick Comments On Time Consistency And MMT

Jackson Mejia and Brian C. Albrecht released a pre-print "Time Inconsistency, Inflation, and MMT" that discusses the one of the theoretical debates about relying on fiscal policy -- the notion of time consistency. This is an old idea, and was part of the neoclassical reaction to Old Keynesian policies. The paper proposes a relatively simple mathematical model, and discusses the implications.


The model follows on from the influential work of Kydland and Prescott published in 1977. The model itself has a number of obvious weaknesses when applied to the Modern Monetary Theory (MMT) framework -- which is discussed in the Mejia and Albrecht paper. However, as they argue, there is a basic principle to be concerned about.

They argue that a similar result will occur so long as two assumptions are true.
First, we make an assumption about preferences: the policymaker is willing to trade off inflation for unemployment. Second, we make an assumption about feasible outcomes: there exists some level of unemployment where a further decrease in unemployment will increase inflation
Although they characterise these assumptions as uncontroversial, one could find objections in the post-Keynesian literature. I would weaken the statements slightly to: policymakers want to trade off short-term growth versus inflation, and excessive short-term growth causes inflation. Although this might appear to be a minor change in wording, it avoids two pitfalls. The first is that it short-circuits any discussion of NAIRU (which is technically not implied by the Maija/Albrecht wording). The second is that if a Job Guarantee is in place, "unemployment" is supposed to be zero.

Sketching the Model

The formal model within the paper is straightforward, although there is a certain amount of tedious algebra in finding solutions. The basic premise is that the government uses lump sum taxes to calibrate growth, trying to lower the unemployment rate while keeping inflation near some optimal level.

The problem within the framework is that there is no way to bind future decisions, and there is an incentive for the central planner to have inflation run persistently higher than optimal.

This model structure follows that of Kydland and Prescott, and runs into a great many objections -- which Mejia and Albrecht address. If we look past the mathematics and just rely on hunches about political economy, one can interpret the 1970s period as a era where politicians were willing to let inflation run higher in the interest of greater growth. (I am agnostic on that view, but having grown up in that era, can see the plausibility behind it.)

The authors argue that some form of commitment strategy is needed, which is in line with the MMT literature.

Complications

The following points can be raised, just working within the current institutional framework.

  • The authors note the Job Guarantee as a stabilisation tool. It is an automatic stabiliser, and so the spending is counter-cyclical. This would not be enough to abolish the business cycle. However, it also means that unemployment technically does not exist. If we assume that the Job Guarantee wage is a living wage, it is unclear that there is a great incentive to drive the number of people in the programme to an extremely low level. As such, other than dealing with recessions, it is unclear how much of a growth/inflation trade-off there is during an expansion.
  • Progressive taxes will act as automatic stabiliser, as is also noted in the article discussion (and cited MMT literature). If wages rise, they will run into higher tax rates, and so the tax take increases without any policy changes. Tax brackets would probably need to move in line with an inflation target.
  • Alternative means of dampening activity can be attempted; e.g., quantitative credit controls. This would give central bank staff something to do while the policy rate is stuck at 0%.
  • Spending would need to be coherent with the policy framework. The tendency to index wages and pensions to inflation would need to be replaced with having them rise in line with the inflation target. Volumes of purchases would need to fall if prices rise.
  • We need to have a plausible model of inflation. Under the current set of labour market institutions -- and not the one of the 1970s -- there is little of evidence of inflation having a tendency to accelerate quickly. Even slow-moving fiscal policy changes may be able to keep inflation as close to target as monetary policy in practice.
In any event, I am not too interested in pursuing technical issues, as we face a much larger point: the model is too simple, and policymakers have too much knowledge of the effects of policy. The current situation is extremely uncertain, but even during normal times, our ability to forecast the growth effects of policy does not allow that much in the way of fine-tuning.

Commitment by Changing the Policy Framework

Debating that particular model is largely beside the point, as noted by Mejia and Albrecht. It seems straightforward to see that different politicians will value the importance of inflation control differently, and so one could imagine that it could be ignored. (In fact, a good portion of critiques of MMT on the internet largely consist of imaginations run wild all the way to Zimbabwe.) 

The obvious answer is to change the budgeting process from worrying about a nebulous "budget constraint" ("what is happening to the debt-to-GDP ratio?") with an inflation constraint. As the title suggests, the article "Replacing the Budget Constraint with an Inflation Constraint" by Scott Fullwiler discusses exactly this topic.

Fullwiler summarises his argument as:
In sum, let’s stop pretending that replacing a budget constraint with an inflation constraint is so hard. It involves a change in perspective, nothing more and nothing less. It doesn’t give license to policy makers to do whatever they want. It does mean CBO will finally be doing something useful with its deficit projections—namely, building models to understand how deficits will affect the macroeconomy (while its current practice is to assume an economy at full employment and warn of impending financial ruin as a result of deficits). 

In the current environment, we face considerable economic uncertainty. Projecting the level of deficits alone is extremely difficult, never mind the effect on inflation. That said, if a Job Guarantee were in place, we would have a better idea where the floor value for nominal income for the household sector is, and so we might have less uncertainty about nominal values.

Returning to the MMT debate, one can argue that the article by Mejia and Albrecht is a step in the right direction. Silly stories about hyperinflation have been dropped, and we need to look at the actual challenges faced by a fiscal-driven policy framework. Since interest rates have been rammed hard against the zero barrier, and there is not a whole lot of enthusiasm for the effectiveness of negative interest rates, the reality is that policy thinking will be forced to move in that direction anyway.

(c) Brian Romanchuk 2020

No comments:

Post a Comment

Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts.

Note: if you want to post comments from Apple devices (iPhone, iPad), you apparently need to turn off "prevent cross-site tracking" in Safari privacy settings. (The reason presumably is that another URL handles comments, and so the user session needs to be preserved when redirected to that site. I don't like this, but this is not enough to make me switch my hosting service.)