86 - William A. Barnett on Divisia Aggregates and Measuring Money in the Economy

Episode Summary

Executive Summary: William Barnett argues that macroeconomics and monetary policy are undermined by faulty monetary measurement. He explains why simple-sum aggregates mismeasure liquidity, defends Divisia indexes as theory-consistent, and shows how broader measures like M3/M4 better captured crisis-era shifts, Volcker-era tightening, and the Great Moderation’s false sense of stability.

Main Topics: Barnett’s path from rocket engineering to macroeconomics (Priority: 3/5): Barnett recounts moving from MIT and Rocketdyne’s Apollo-era engineering work into economics via graduate study, ultimately landing at the Federal Reserve Board, where his statistics and finance background positioned him to rethink money measurement. Why simple-sum monetary aggregates are flawed (Priority: 5/5): He argues that adding monetary assets equally assumes perfect substitutability, which is inconsistent with user-cost theory and modern financial reality. Simple-sum measures treat very different assets as if they provide identical liquidity services. What Divisia aggregation fixes (Priority: 5/5): Barnett explains that Divisia indexes use user-cost prices and aggregation theory to weight monetary components according to their liquidity services, making them more consistent with microeconomic theory and empirically more stable. M1, M2, M3, and L/M4 compared (Priority: 4/5): He walks through the composition of standard aggregates and highlights anomalies: M2 includes illiquid non-negotiable CDs but excludes liquid negotiable CDs; M3/L (now M4) capture wholesale liquidity and shadow-banking instruments more relevant to large firms. The Barnett critique and money-demand stability (Priority: 5/5): The Barnett critique says apparent instability in money demand can arise from inconsistent aggregation methods. If the aggregate is mismeasured, the estimated demand function may look unstable even when the underlying behavior is stable. Crisis-era and policy implications (Priority: 5/5): Barnett links mismeasurement to the Great Moderation, Volcker disinflation, and the 2008 crisis, arguing that broader Divisia measures would have better signaled tightening, liquidity shortages, and excessive confidence in the Fed’s abilities. Institutional and policy reform (Priority: 4/5): He notes that several central banks use Divisia aggregates internally or publicly, and advocates a dedicated Bureau of Financial Statistics within the Fed, akin to the BLS or BEA, to improve monetary data quality.

Key Arguments: Simple-sum money aggregates are theoretically inconsistent because they assume equal weighting of assets with very different liquidity services. Divisia indexes are superior because they are derived from index-number and aggregation theory and weight assets by user-cost prices. M2 contains paradoxical component choices, such as illiquid non-negotiable CDs while excluding more liquid negotiable CDs. M3/L (now M4) better capture the money relevant to large corporations and the wholesale/shadow banking system. Apparent instability in money demand can be a measurement problem rather than evidence that money demand is inherently unstable. The Great Moderation fostered excessive confidence in central banks partly because standard monetary measures failed to reveal underlying liquidity shifts. Volcker genuinely targeted money growth, but simple-sum measures understated how much tighter Divisia measures indicated policy had become. A specialized data bureau inside the Federal Reserve would improve monetary measurement and align it with best practice used in other statistical agencies.

Data Points: Rocketdyne engine thrust: 1.5 million pounds of thrust - Barnett describes the Apollo-era F1 rocket engine he worked around, emphasizing the scale of the engineering environment. Educational leaves at Rocketdyne: 1 year of educational leave per year employed - He says Rocketdyne gave generous benefits that allowed him to pursue graduate study in finance, economics, and statistics. NASA contracts: Cost-plus fixed fee - Barnett uses this to illustrate weak incentives in early space-program contracting. Years of major funding shift: Vietnam War era - He explains that NASA funding declined while Defense Department funding rose, altering his career path. Volcker-era monetary growth comparison: Divisia aggregates grew about half as fast as simple-sum aggregates - Barnett reports this from his presentation to the Federal Reserve Board during the monetarist experiment. Federal Reserve discontinued tracking: M3 in 2006 - He notes the Fed stopped publishing M3 and earlier stopped L/M4, which he considers a major data loss. Great Moderation period: 1980s-2007 (implied period) - Discussed as a time when central banks appeared more effective than they really were, in Barnett’s view. Central bank use of Divisia: Bank of England, ECB, Bank of Japan, Bank of Poland, Bank of Israel, IMF - Barnett says several institutions already use or advocate Divisia measures, though not always publicly.

Pivotal Quotes: "the simple sum aggregates are very simply incompetent" — William Barnett: His direct critique of conventional monetary aggregation methods using equal-weight summation. "the demand for money function is surprisingly stable" — William Barnett: He argues that stable money demand is visible when money is measured properly with Divisia indexes. "if I had just told me you want to produce the Visi indexes with user cost prices, I would have agreed right away" — Gary (as recalled by Barnett): Barnett recounts how economists quickly grasped the logic once the measurement approach was framed in standard aggregation terms.

Implications: For analysts, the transcript argues that monetary policy should be judged with better liquidity measures, especially Divisia M3/M4. For institutions, it supports building stronger statistical infrastructure so hidden financial stress and policy errors are easier to detect early.