Is America Getting Less Growth From More Debt?

When we think about debt in an economy, the central question isn’t just how much debt we have, but rather how productive that debt actually is. To answer this, we looked at the Marginal Productivity of Debt (MPD).

What is MPD?

Marginal Productivity of Debt measures how many dollars of incremental economic growth (i.e., Gross Domestic Product, or GDP) are generated for every new dollar of net credit added to the economy over a given period. Mathematically, MPD is represented as follows:

Change in Nominal GDP ÷ Change in Net Debt

What this measure tells us: If, for example, MPD is +$0.75, it means that for every $1.00 of new borrowing, our annual economic output increased by $0.75. If, let’s say, MPD drops to +$0.35, the economy requires almost $3.00 of new borrowing just to generate that same $1.00 of growth (simply divide 1 over 0.35 to observe the precise value). As we can see, MPD is a direct temperature check, if you will, on the “bang for the buck” of credit expansion.

Data and Methodology

To build a clear, historical picture from 1952 through 2026, we pulled macroeconomic data directly from our favorite tool, the Federal Reserve Economic Data (FRED) database from the St. Louis Fed:

1. Economic Output Data Series (Numerator): “GDP” (Gross Domestic Product, Quarterly, Billions of Dollars)

2. Total Nonfinancial Debt Data Series (Denominator): “TCMDODNS” (Domestic Nonfinancial Sectors; Debt Securities and Loans; Liability, Level, Millions of U.S. Dollars). This captures borrowing by households, nonfinancial businesses, and federal/state governments, while stripping out financial sector intra-bank transactions to avoid double-counting.

3. Data Transformations Applied:

• Evaluated the 1-year trailing change (“change from Year Ago”) for both GDP and Debt to capture the rolling 12-month net expansion.
• Harmonized units by scaling debt from millions into billions using the formula: (∆GDP x 1000) ÷ ∆TCMDODNS 

Key Observations from 70+ Years of Data (1952-2026)

Let’s examine what the graph we generated shows (red circle and lines on the graph are my markings):

The 70-year shift in U.S. debt efficiency is stark, and the core takeaways from this analysis are as follows:

1950s – 1970s: High Efficiency Era (range from ~+$0.60 to ~+$0.85 per $1.00 of debt):

• Debt directly funded industrial capex, factory equipment, and physical infrastructure.

1980s – 2026: Financialization & Diminishing Returns Era (range from ~+$0.25 to ~+$0.40 per $1.00 of debt):

• Shift toward asset refinancing, real estate turnover, and servicing legacy debt overhang. 

1. The Structural Decline in Debt Efficiency is Real

Prior to the early 1980s, the U.S. economy routinely generated ~+$0.60 to ~+$0.85 in annual GDP for every $1.00 of net new debt issued outside of recessions. Starting in the early 1980s, this baseline experienced a persistent structural downward shift. For the past two decades (2000s through 2026), the baseline has hovered between ~+$0.25 to ~+$0.40. It now takes significantly more credit to generate the same unit of economic expansion.

2. Diminishing Marginal Returns to Credit

Why did this happen? Here we draw from our experience and understanding to come up with the likely culprits.

From Productive Capex to Asset Refinancing: In the post-war era, debt primarily funded tangible, capacity-expanding capital investments (e.g., manufacturing plants, highways, utility grids). Over the last four decades, an increasing share of borrowing shifted into real estate asset turnover, corporate financial engineering (e.g., share buybacks and M&A), and government deficit spending.

The Drag of Debt Overhang: As total accumulated debt expanded relative to GDP, a growing share of every newly borrowed dollar had to be spent simply servicing, rolling over, and refinancing existing debt rather than building new productive capacity.

3. Debt Does Not Prevent Recessions (Credit Cycle Vulnerabilities)

I like to think of this part as a double-edge sword, because debt is a powerful tool to amplify returns both on the way up and on the way down. You will see in a moment what I mean. Indeed, while credit is a powerful tool at pulling future demand into the present during expansionary periods, it does not immunize the economy from downturns:

• In every major recession – 1954, 1958, 1974, 1982, 1990, 2001, 2008, and 2020 – the MPD dropped sharply, dipping deep into negative territory (reaching ~-$0.40 during the Great Financial Crisis).
• Because debt creates fixed payment obligations while economic income fluctuates, high leverage actually amplifies downturns. When income contracts, borrowers are forced to slash spending, payroll, and investment to service their debt, turning mild slowdowns into sharper contractions. On this point, I think the so-called Austrian School of Economics has it mostly right in their view of the economic cycle, but that is a story for another day.

Conclusion:

Debt is a critical tool for economic growth, but it has its limits. The historical data from 1952 to 2026 shows that the U.S. economy is getting diminishing marginal output from each new dollar of credit, and accumulating more debt cannot eliminate the cyclical reality of recessions. Unless something structural occurs in the economy that improves the marginal output from each new dollar of credit, the U.S. will at some point find itself facing serious economic issues. We cannot indefinitely borrow ourselves to prosperity.