The craziest year in recent memory has illuminated an emerging macroeconomic trend amongst the world’s financial circles. Rapid expansion of the fiat money supply has led to fiery arguments on both sides of the aisle; some are well-formed and some lack substance. Proponents believe governments have a moral duty to use the money supply to correct undesirable current economic metrics, while others argue that aggressive money printing will lead to pain in some way or another. As a former physics student who is well aware of the infeasibility of perpetual motion machines, much of my personal worldview is shaped by the latter analysis- however, I felt that it would be intellectually dishonest to support that opinion without fully understanding the argument for the other side. My pursuit of objectivity led to Stephanie Kelton’s The Deficit Myth, a book that illustrates the philosophy of Modern Monetary Theory and advocates for its (continued) use in today’s political system.
MMT is centered around the idea that a goverment’s money printing ability relieves budgetary pressure to pay back debt. Kelton notes that the limit to a government’s financing power should be the point where inflation grows beyond an undesirable level. Of course, the inflation constraint is of paramount interest here. Heading into the book I was hoping for a few core questions to be answered:
- How can an MMT-based government evaluate a given’s policy’s impact on inflation?
- With what certainty can we predict inflation?
- What tools can MMT give us to curb inflation if/when it grows?
While I respect Kelton’s conviction in a new school of thought, her work has left me unsatisfied. To be frank, inflation seemed like a footnote to Kelton’s main point, which was that much of today’s current monetary thought revolves around the idea of maintaining a balanced budget. The Deficit Myth argues that sovereign countries can expand the money supply to meet debt obligations, which is categorically true. But the question is, should they?
My understanding is that Kelton believes an MMT-based government (specifically, a central bank) has the ability to answer the above questions with historical data and empirical evidence. My main goal with this post is bring risks, ethics, and incentives of a MMT-based economy into question.
Several lessons from HFT apply to the MMT rationale. Often as a researcher in quantitative finance, the objective is to tune an algorithm’s parameters such that future trades will capture more upside and less downside. One approach is to employ a technique called backtesting, which entails simulating a strategy’s performance by advancing through historical data and observing hypothetical performance vs some benchmark. While clean on the surface, backtesting often ends up as a stab in the back of a trading operation. Backtesting can be flawed as an alpha-generating approach for several reasons, but the main culprit is overfitting. Unfortunately, market-based systems evolve over time and go through regime changes- stochastic processes are not good candidates for backtesting because there is no guarantee that observed patterns in the past continue on into the future.
Going back into the monetary policy conversation, it should raise questions if we are using historical data to evaluate whether prices of goods will increase or decrease given a change in tuning parameters. We have roughly 50 years of data since Bretton Woods was dissolved in 1971- a tiny sample size in the scope of human civilization. The onus is on the researcher, not the critic, to evaluate the accuracy, risk, and applicability of market models.
If the argument for justifying MMT is that money printing has not yet led to inflation, what fundamental principle asserts that this historical performance should indicate future performance? Taleb’s turkey would like to have a word.
You may have heard of John Locke’s “social contract” in school, which describes every member of society’s implicit obligation with another to guarantee fundamental human rights. On an abstract level, you could think of this contract as existing along a socioeconomic axis. If we advance along an axis of socioeconomic status, we would expect the axioms of the contract to exhibit little variance; in other words, individuals that vary in socioeconomic status should not vary in terms of rights endowed by the social contract.
Rarely do I see the idea discussed of a social contract that exists along a time axis. What I mean by that is:
Do we, as individuals living in the society of 2020, have a social contract with the individuals that live(d) before and after us?
I’d argue that the public opinion does subscribe to this time-oriented social contract, as concerns over climate change seem to indicate that we do hold an obligation to future civilizations (or, at least to some degree).
Reconcile the time-oriented social contract with Kelton’s argument for printing money in order to fund things like social security and a federal job guarantee. Most would agree that retirement stability and low unemployment are good outcomes, but at what cost? Should we execute these policies if they imply a 1% probability that our society will suffer 50 years down the line due to monetary expansion? What if that probability was 20%, or 95%? If a policy decreases unemployment by 10% today and increases it by 11% 10 years from now, is that a good trade?
A pessimist on human nature would assert that the political incentives to trade short-term problems for long-term ones are quite enticing.
The strongest argument for MMT would establish an acceptable baseline for expected inflation-driven damage to society. By conceding that money printing must not induce significant inflation, MMT concedes that every dollar printed produces inflation with some probability.
Inflation should be a controversial topic in itself. Firstly, how is it defined? A general definition would include something to the effect of “a reduction in the general purchasing power of money”. An economist might be slightly more specific, stating that inflation is the change in Consumer Price Index year-over-year. From an individual’s perspective, CPI only tells a fraction of the story. How much of an individual’s income is devoted towards goods in the CPI? Of course, this proportion varies from person-to-persion, but I would argue that in aggregate, this proportion is low- certainly far lower than 100%.
For example, a resonable goal for an average person in the U.S. may be to acquire a house at some point. If we are to measure this person’s “purchasing power”, why are real estate prices not accounted for? Let’s say for a given year, a person spends 50% of income on housing, CPI is unchanged, and housing prices rise by 20%. The government would tell us that inflation is zero, when in reality, this person’s purchasing power has eroded by
.50 * (1 - .80) = 10%. If someone with a post-tax income of 50k USD/year dreams of buying a house priced 500k USD today, that person is treading water if the price of housing accelerates at a rate >= 10% faster than income. In reality, this threshold is even lower, due to compound returns and static marginal tax rates.
If we expand this idea to other assets such as stocks, bonds, and commodities, it’s easy to see why CPI doesn’t tell the whole story in terms of general purchasing power.
This isn’t to say that the CPI conveys no information; humanity would certainly prefer that the price of Apple stock goes up 10,000% than the price of water go up by 10,000%. But to say that asset inflation is not real inflation is an unrealistic assertion that conveniently gives the government more liberties in terms of monetary manipulation. The reality is that everyday people spend money on goods beyond those classified under “everyday consumption”, and when those goods rise in price, purchasing power falls.
The Inflation Myth is that the central bank view of inflation describes purchasing power.
Income inequality troubles most people today, including MMT advocates like Kelton. Unfortunately, income inequality is a negative externality of a capitalistic society, and usually requires government intervention to correct itself.
Ironically, MMT drives income inequality far more than any flawed tax legislation. The excessive money printing endorsed by MMT accelerates wealth growth for those that already have the capital to capture asset inflation. As long as USD holders seek refuge in assets, this trend will continue. Cantillion has shown us that money injection often does not provide relief to the areas of the population that need it the most.
I do believe that MMT will grow in popularity if not for purely political reasons. Being able to solve our short term problems with no apparent costs appears irresistible; ignorance is bliss.
The spigot has been opened. Do we have an implicit responsibility to prevent a flood?