Why Study Nearly a Century of Money Data? A First Reading of A Monetary History

Why Study Nearly a Century of Money Data? A First Reading of A Monetary History

A Monetary History of the United States begins in 1867 not because Friedman and Schwartz wanted a neat century-long span, but because they considered it the earliest year from which a continuous U.S. money-stock series could be estimated. The National Banking Act disrupted federal statistical collection for state banks, leaving a serious gap in the 1863–1867 record. That choice reveals the method behind Chapter 1: define the monetary measure, reconstruct the data and institutional setting, and examine how money and economic activity influenced each other.

The opening sentence is direct: “This book is about the stock of money in the United States.” Chapter 1 The authors make clear that this is not a comprehensive history of the American economy. The book discusses banking, politics, wars, the gold standard, and business cycles, but it follows one thread: the stock of money. Which episodes receive detailed treatment—and which are mentioned only briefly—depends on their relevance to that thread. This restriction gives the book a clear analytical boundary. Rather than explaining everything, it repeatedly asks how one observable monetary measure changes under different institutions and historical conditions.

Why begin in 1867?

The starting date is not a convenient round number. Friedman and Schwartz explain that the National Banking Act passed during the Civil War disrupted the federal collection of statistics for state banks. State banks did not disappear as some policymakers had expected, but the statistical record developed a serious gap. For the purpose of constructing a continuous money-stock series, the years from 1863 to 1867 were not sufficiently reliable. The year 1867 was the earliest point from which the authors believed they could begin a continuous estimate.

This is a useful reminder about historical research: the subject of a study is not determined by theory alone. Data availability, continuity, definitions, and institutional changes also determine where a reliable comparison can begin. A chart extending from 1867 to 1960 may look like a uniform modern database, but it is not. Anna Schwartz later explained that the frequency of the early estimates varied: annual estimates before 1875, semiannual estimates from 1875 through 1880, annual estimates again from 1881 through 1906, and continuous monthly estimates beginning in May 1907. A long historical series is valuable, but its observations do not all have the same frequency or precision. Schwartz described this sequence in a 1993 interview.

What did the authors mean by “money”?

The introduction also makes an important point about definition. Under the measure they used for 1867, money included currency and commercial-bank deposits. The authors estimated approximately $585 million in currency and$729 million in commercial-bank deposits, for a total of about $1.314 billion [in Chapter 1](https://www.jstor.org/stable/j.ctt7s1vp.4). If deposits at mutual savings banks were included under a broader definition, another$276 million would be added, bringing the broader total to approximately $1.590 billion.

Why did the authors not simply separate demand deposits from time deposits? In the late nineteenth century, the distinction did not have the same practical meaning it acquired later. Time deposits could pay interest and sometimes be transferred by check; demand deposits could also carry interest arrangements. The distinction became more important to banks after the Federal Reserve Act, enacted in 1913, established different reserve requirements for demand and time deposits and the Federal Reserve System began operating in 1914; reliable continuous data for the two categories then became available. The authors therefore could not construct a narrow 1867 measure equivalent to currency plus demand deposits alone.

This matters because the historical money stock in the book should not be casually relabeled as today’s M1 or M2. Monetary definitions change as banking institutions, financial products, and statistical systems change. A series may appear continuous while the economic object being measured evolves. Before asking how much money grew, a careful reader should first ask what the measure includes.

What a long historical series makes possible

From 1867 to 1960, the United States experienced very different monetary arrangements and financial structures. The post-Civil War greenback period, the resumption of gold payments in 1879, late-nineteenth-century silver politics and changes in gold supply, the passage of the Federal Reserve Act in 1913 and the Federal Reserve System’s start of operations in 1914, the expansion of the System in the 1920s, the Great Contraction, New Deal banking reforms, wartime inflation, and postwar monetary developments all appear within the same historical framework.

The value of this comparison is not that every period produces the same conclusion. It is that similar-looking economic developments can operate through different mechanisms under different institutional arrangements. Changes in the money stock may reflect real output, prices, the public’s willingness to hold money, banking structure, policy decisions, or the international monetary system. Without the institutional context, a single growth rate can easily be mistaken for a complete explanation. Historical analogy can then become a substitute for evidence. For a broader view of financial-market cycles, see The Great Game Revisited: Wall Street History, AI Bubbles, and Digital Assets; it looks at capital-market narratives rather than the money stock itself.

The authors report in Chapter 1 that, by their chosen definition, the U.S. money stock increased about 157-fold between 1867 and 1960, an average annual rate of roughly 5.4 percent. Adjusted for population growth, money per capita increased about 32-fold, or roughly 3.7 percent per year. They then decompose the per-capita increase into growth in real output per person, rising prices, and an increase in the amount of money balances the public chose to hold relative to income. The important point is that growth in the money stock is not synonymous with inflation. Money, real output, the price level, and the public’s desired money holdings are connected variables, but they are not interchangeable.

Money and economic activity do not move in only one direction

Another important feature of the book’s method is that the authors do not treat money as a completely independent dial controlled only by the central bank. In the introduction, they state that monetary factors played a major role in business-cycle movements, while nonmonetary developments frequently had a major influence on monetary developments. Economic activity, financial structure, the public’s choice between currency and deposits, bank balance sheets, and central-bank policy can all affect the money stock at different times.

The authors use the National Bureau of Economic Research’s reference chronology of business-cycle peaks and troughs to compare monetary changes with economic activity. They also caution that the book is not a comprehensive history of U.S. business cycles. This is important: a long-run narrative can reveal recurring relationships, but it does not automatically prove that one variable caused every movement in another. A downturn can affect the public’s demand for money; banking stress can then amplify the downturn by changing deposits and credit. The relationship may be a feedback process, not a one-way arrow.

The book is therefore both a work of economic theory and a work of economic history. In his 1989 review, Michael Bordo described its two legacies: using history to develop the modern quantity theory of money, and reinterpreting U.S. monetary history through the relationship between money and the wider economy. Subsequent researchers have extended and criticized the authors’ interpretations of individual episodes. A classic work does not end debate; it creates a framework clear enough for later scholars to test, challenge, and refine.

The framework itself is not beyond debate. In their 1989 paper, “Does Monetary Policy Matter? A New Test in the Spirit of Friedman and Schwartz”, Christina and David Romer explicitly took the narrative approach pioneered by Friedman and Schwartz as a starting point: they used historical records to identify monetary disturbances not caused by output fluctuations and test their real effects. The paper’s central analysis examines six postwar U.S. episodes in which the Federal Reserve shifted toward contraction to reduce inflation, while also discussing interwar evidence. It did not simply reject narrative monetary history; it made the identification of independent shocks a testable question. In a 1993 interview, Schwartz explained that the Romers questioned whether the shocks selected in A Monetary History were biased; in their view, only the discount-rate hikes in 1920 and 1931 qualified as monetary disturbances. She also noted that they did not believe the possible selection bias erased the contractionary effects of monetary shocks. In 2013, Romer and Romer returned to a different but related issue: the transmission mechanism in the Great Depression explanation. Observing falling nominal interest rates was not enough; the monetary explanation required evidence that deflation expectations were caused by monetary contraction. Their analysis of Business Week found evidence that monetary contraction and Federal Reserve policy contributed to deflation expectations, which may have depressed spending and output partly by raising real interest rates. The debate thus moved from selecting historical episodes to identifying independent shocks and tracing how they affected the real economy.

Anna Schwartz was not a secondary name on the cover

The authors were Milton Friedman and Anna J. Schwartz. Public discussion often makes Friedman the central figure, but Schwartz was not a nominal co-author. In a 1993 interview, she explained that the project grew out of the National Bureau’s effort to study cyclical changes in different economic processes. The authors had constructed estimates of the U.S. money stock from 1867 to 1960 and were assigned to study the cyclical behavior of those estimates: what accounted for fluctuations in money growth, and how those changes affected economic activity.

A 2024 archival study in the Journal of Economic Perspectives further identified Schwartz’s contributions to the collaboration: collecting and assessing historical archival data, shaping the statistical approach, and helping turn the material into a persuasive historical narrative. This matters for more than correcting the byline. It shows what serious economic history requires: long-term data collection, statistical judgment, interpretation of historical evidence, and the ability to connect those materials into an argument that can be examined by others.

What should readers take from the book—and what should they not copy?

The financial system of 1960 is very different from that of 1867, and both are different from today’s. Bank accounts, financial intermediaries, payment instruments, capital markets, and central-bank operating frameworks have all changed. In a 2006 speech, then-Federal Reserve Chair Ben Bernanke noted that financial innovation and regulatory changes had repeatedly destabilized the relationship between monetary aggregates and other nominal variables. Since 1982, monetary and credit aggregates had not played a central role in U.S. monetary-policy formulation, although policymakers continued to use monetary data as information about the state of the economy. This is a historical reminder about changing measurement and institutions, not a claim that money data have become irrelevant.

The lesson is not to turn the book into a simple investment formula: buy when money growth accelerates and sell when it slows. The more useful lesson is how to ask questions, build a dataset, identify institutional changes, and test relationships among variables. Historical data can show how a mechanism operated in one setting; they cannot guarantee that the same mechanism will operate with the same strength, speed, or direction today.

For investment research, the method can be turned into a short macro-data checklist. First, record the definition of the measure: money stock, bank credit, the monetary base, or the central bank’s balance sheet? Second, record the data basis and frequency: nominal or real, monthly or annual, and whether historical definitions have been revised. Third, identify the institutional regime: gold standard, fixed exchange rate, or modern fiat system; what was the structure of the banking system and the central bank’s operating framework? Fourth, distinguish possible directions of transmission: did money affect income, did income affect money demand, or did both reinforce each other? Fifth, write down what evidence would falsify the current explanation rather than keeping only the observations that support it.

My reading is that the book’s most durable contribution is not a monetarist trading rule but a sequence for checking macroeconomic claims: define the measure, reconstruct the institutional setting, and test the transmission mechanism. This checklist cannot identify which stock to buy, and it cannot replace financial statements, valuation, or industry research. Its purpose is to prevent a single macroeconomic number from being treated as a complete explanation. When money aggregates rise, we still need to ask whether credit reaches banks, businesses, and actual spending. When nominal rates fall, we still need to examine credit availability, spreads, and corporate cash flow. Historical research improves the quality of the questions; it does not remove the need for judgment. For a current example of separating aggregate growth from household experience, see Why Is GDP Still Growing While Everyday Life Feels Colder?.

The first lesson of A Monetary History

The book begins in 1867 not because its authors wanted a neat century-long story, but because that was the earliest point at which they could construct a continuous estimate of the money stock. It does not attempt to cover every aspect of U.S. economic history. Instead, it follows one monetary thread through different institutions and business cycles, examining how the stock of money changed and how it interacted with economic activity. Its contribution is not only an explanation of a few crises. It is a method: define the variable, inspect the data, reconstruct the institutional setting, consider causality in both directions, and leave room for later research to test the conclusion.

For today’s investor, the most useful takeaway is not a label from one school of thought, but a disciplined skepticism toward macroeconomic data. Before asking what a number means, establish what it measures. Before using a historical case as an analogy, establish how similar the institutional conditions really are. Reading history should not make us faster at reaching conclusions; it should make us more accurate in framing the questions that come before them.

The point of studying the past is not to force it onto the present, but to identify which conditions have changed and which mechanisms still deserve to be tested.

If you want to turn reading into a repeatable decision process, see I No Longer Write Book Reviews: How a Programmer Turns Reading into a Decision System.

Edition, sources, and scope

This note focuses on Chapter 1, “Introduction,” of Milton Friedman and Anna J. Schwartz’s A Monetary History of the United States, 1867–1960 (Princeton University Press, 1963), supplemented by Romer and Romer’s 1989 methodological study, Schwartz’s 1993 interview, Michael Bordo’s 1989 scholarly review, Jennifer Burns’s 2024 archival study, Ben Bernanke’s 2006 speech on monetary aggregates, and Romer and Romer’s 2013 study of the monetary transmission mechanism during the Great Depression. It is a reading note on the book’s research method, not a claim to have completed a chapter-by-chapter reading of the entire volume. Historical figures retain the definitions used by the authors and should not be treated as modern M1, M2, or national-accounting measures.

1. Friedman & Schwartz, A Monetary History of the United States, 1867–1960 (NBER book record)
2. Chapter 1, Introduction (JSTOR book and chapter record)
3. Anna J. Schwartz interview: project origins, data frequency, and collaboration (Federal Reserve Bank of Minneapolis)
4. Michael D. Bordo, The Contribution of a Monetary History of the United States: 1867 to 1960 to Monetary History (NBER, 1989)
5. Christina D. Romer & David H. Romer, Does Monetary Policy Matter? A New Test in the Spirit of Friedman and Schwartz (NBER Macroeconomics Annual, 1989)
6. Jennifer Burns, Retrospectives: Friedman and Schwartz, Disaggregated (Journal of Economic Perspectives, 2024)
7. Ben S. Bernanke, Monetary Aggregates and Monetary Policy at the Federal Reserve: A Historical Perspective (Federal Reserve, 2006)
8. Christina D. Romer and David H. Romer, The Missing Transmission Mechanism in the Monetary Explanation of the Great Depression (American Economic Review, 2013)