.png)
May 6, 2021 | In this issue of The Institutional Risk Analyst, Brian Barnier (www.feddashboard.com and City University of New York) discusses why the "data dependent" analytical process used by the Federal Open Market Committee for measuring inflation may be measuring the wrong thing.
Price indices don’t really say what policymakers think they say.
Central bankers often act as though price index statistics are the same as the “generalized price level” of theory -- and the level is caused primarily by monetary factors. Yet, hard data and daily life – especially during COVID – show that this assumption is false. That means a 2% target of any flavor isn’t what the FOMC thinks it is. The Fed's 2% inflation target needs to be replaced in a specific way. And investors need to embrace the data in order for the measure to be credible.
“Monetarism” is widely debated. Oddly, the monetarist assumption of “inflation” is widely accepted. “Inflation” is assumed to be 1) change in generalized price level that is primarily caused by money supply (or at least monetary factors) and 2) that the “generalized price level” is statistically represented by a weighted aggregate average of product prices.
The Federal Open Market Committee looks to the Bureau of Economic Analysis’ (BEA) Personal Consumption Expenditures (PCE) Implicit Price Deflator. (Statistically more appropriate for monetary policy than the Consumer Price Index (CPI) used for Treasury Inflation-Protected Securities.) This index is calculated from hundreds of product prices obtained or imputed by the BEA, Bureau of Labor Statistics (BLS) and Census Bureau.
The Five Realities
In a “Stats 101” sense, price indices are invalid for statistical inference because the average lacks a single mode and stable standard deviation. This is because prices by category have been diverging. Services prices up, nondurable goods (think food and energy) flat for a decade and durable goods down since 1995 (starting with electronics in the 1970s).
Price categories changing in different directions – not a uniform monetary cause, as illustrated in the chart below.
Falling durable goods prices are mostly the result of the global tech and trade transformation (with a recent uptick from the Trump-Biden tariffs and COVID). Plus, shoppers buy more when prices fall – busting the myths that falling prices mean weak demand and rising prices are needed fear to “induce” purchases.
Rising services prices have mostly come from housing (it is estimated as a service of imputed rent), health care and financial services.
Services are driven by public policy. Regarding health care, then Federal Reserve Chair Janet Yellen addressed health care policy in her 2017 speech to the National Association for Business Economics. Housing is local. Financial Services & Insurance (4th largest category in the PCE deflator) has long been a squishy number, not updated for the past decades of changes in financial markets and fintech.
Absolute levels and change rates vary widely geographically as seen in the BLS CPI local data. (Kudos to the St. Louis Federal Reserve Bank’s FRED team for making this easy to see.)
The rationale for “Core PCE” that subtracts product price causes beyond monetary causes could also remove most items in the PCE deflator. The rationale of agricultural technology after WWII or geopolitics for energy extends to most durables and services.
There is little product-level evidence for monetary policy impact on prices. There are cases – such as the Effective Federal Funds Rate on house prices or foreign exchange rates that cascade to import prices. Yet, a broad mechanism is elusive.
Central bankers can ask themselves some basic questions:
1. Is the cause of change in a product price due to monetary policy?
2. If not, is it a central bank’s role to counter nonmonetary causes? For example, upcoming stimulus and infrastructure laws, Trump-Biden tariffs, USMCA trade agreement, health care laws, people relocating due to COVID or to cut their taxes, preferences for luxury or convenience, falling costs of manufacturing, global supply chain reorganization (due to COVID or geopolitics) or the COVID spending skew.
3. What are the complications of using monetary policy tools in response to a nonmonetary cause of inflation? Will central banks set a threshold of acceptable distortions and inequities?
Importantly, these questions aren’t about transitory verses longer product price increases, or asset prices increases. They are about cause. The problem of distinguishing between monetary and nonmonetary causes of inflation isn’t new.
Official statistics, for example, show this situation is long-growing – starting in the 1970s as consumer electronics prices began to fall. Falling consumer electronics and oil prices did much to help Federal Reserve Chair Paul Volcker “break the back of inflation” in 1981.
In official statistics and industry data, it’s been difficult to find theoretical “inflation” in open, industrial economies for decades. Even in the U.S. in the 1970s, the oil embargo, rise of the credit card, end of the Bretton Woods system and more makes identifying monetary causes difficult.
Why are central banks so slow to adjust? Is it a structural problem?
Back to the Future
Irving Fisher, in his Purchasing Power of Money (1911), showcased the mechanics of how things work. He famously diagramed a merchant’s scale to illustrate the role of money and listed causes of product price change. His list stands the test of time (he missed health care).
John Maynard Keynes in his General Theory (1936) focuses us on data and offers us encouragement to change. First, in data. Keynes wrote of Alfred Marshall:
“It seems that we have been living all these years on a generalization which held good, by exception, in the years 1880−86, which was the formative period in Marshall’s thought in this matter, but has never once held good in the fifty years since he crystallised it!”
Today, it is nearly fifty years since the post-WWII period that enshrined current views of “inflation.” Second, Keynes concluded his preface with the following warning to future generations of economists:
“— a struggle of escape from habitual modes of thought and expression. The ideas which are here expressed so laboriously are extremely simple and should be obvious. The difficulty lies, not in the new ideas, but in escaping from the old ones, which ramify, for those brought up as most of us have been, into every corner of our minds.”
The FOMC makes a great fuss about embracing hard data and mechanics. Perhaps central banks would better serve consumers, businesses and investors by embracing a more reliable measure of inflation. A leading candidate is the stability of the trend in consumer durables, after a border adjustment (for trade barriers and FX). But as Keynes said nearly a century ago, the first task is letting go of obsolete and outmoded measures of price changes.
