What is Driving the Recent Rise in the Natural Rate of Interest?

Structural drivers of the U.S. natural rate

Carlos Carvalho, Andrea Ferrero, Felipe Mazin, and Fernanda Nechio, in this paper, use an error correction model to estimate long-term relationships between real (inflation-adjusted) interest rates and their underlying determinants across a panel of advanced open economies. The authors analyze annual data from 1979 to 2019 for 19 advanced economies, including the U.S., major European countries, Japan, Canada, Australia, and others. The model examines the impact of:

1. Demographic trends (population aging, life expectancy, population growth)
2. Productivity growth (Total Factor Productivity, TFP)
3. Fiscal conditions (government spending and debt)
4. International/global factors (captured by a “global rate” reflecting average real rates across countries)

The analysis incorporates international linkages, recognizing that savings and investments flow across borders. The model uses a weighting scheme to account for each country’s financial integration and includes a global rate to proxy for international trends. The estimated relationships are used to decompose the contributions of each factor to changes in the U.S. real interest rate, both before and since the COVID-19 pandemic.

Key findings:

• Pre-Pandemic Period (1990–2019): The persistent decline in U.S. real interest rates was largely driven by global factors and U.S. population aging. International trends (as captured by the global rate) and domestic demographic shifts exerted significant downward pressure on real rates. Fiscal spending provided some upward pressure but was not enough to offset the downward trend. Productivity growth (TFP) played a minimal role in driving the neutral rate.
• Post-Pandemic Period (Since 2019): The downward pressure from global factors and U.S. population aging has faded. Fiscal conditions (increased government spending and debt) are now putting more upward pressure on U.S. real interest rates. International factors have shifted from being a source of downward pressure to exerting upward pressure on U.S. rates. The contribution of population aging to lower rates has diminished, while fiscal factors have become more prominent. TFP continues to have a negligible effect on the neutral rate.

Underlying Trends in the U.S. Neutral Interest Rate
Authors: Carlos Carvalho, Andrea Ferrero, Felipe Mazin, Fernanda Nechio
From: PUC-Rio, University of Oxford, Penn Wharton Budget Model, Federal Reserve Bank of San Francisco

Survey-based estimates of the natural rate

Jonathan S. Hartley aims to provide new, atheoretical (model-free) estimates of the natural rate of interest (r-star, or r*) by leveraging survey-based measures of market participants’ expectations. This approach is contrasted with traditional model-based (structural) estimates, as in the paper above,  which are susceptible to model misspecification and often have large standard errors. Hartley uses long-run policy rate expectations from central bank surveys conducted in the US, UK, Eurozone, and Canada.

Key findings:

• Divergence since the Pandemic:
  – Both survey-based and structural r-star estimates declined before the COVID-19 pandemic.
  – However, in the early 2020s, survey-based r-star estimates have risen, while structural estimates have continued to fall, particularly during the period of resurgent inflation.
  – This divergence is observed across all surveyed regions (US, UK, Eurozone, and Canada).
• Survey estimates track central bank projections:
  In the US, survey-based r-star estimates closely track the Federal Open Market Committee’s (FOMC) Summary of Economic Projections (SEP) for the longer-run federal funds rate, with differences rarely exceeding 0.13 percentage points since 2012.
• Policy implications (Taylor Rule Analysis):
  – Using survey-based r-star estimates in Taylor Rule calculations results in higher recommended policy rates during the early 2020s than using structural r-star estimates.
  – Inertial Taylor Rule variants (which smooth policy changes over time) also show that survey-based r-star estimates support a less aggressive easing or tightening cycle than structural estimates.
• Advantages of survey-based estimates:
  – Survey-based estimates are not subject to the same model misspecification risks as structural models and do not embed risk or term premia issues present in market-based measures (e.g., long-term bonds).
  – These estimates are gaining traction among policymakers, with some Federal Reserve officials explicitly considering both survey and model-based r-star measures in their deliberations.

Survey Measures of the Natural Rate of Interest
Author: Jonathan S. Hartley
From: Stanford University

Comparison of two papers

Complementary insights

• Hartley highlights the value of real-time expectations for policymakers, avoiding model misspecification risks. Surveys align closely with central bank projections (e.g., FOMC SEP).
• Carvalho et al. emphasize structural drivers: demographics and fiscal policy are critical for long-term r-star trends, but global integration complicates domestic estimates.

Divergences

• Post-pandemic r-star: Hartley’s survey data suggests a rebound, while Carvalho et al. point to the shift to fading demographic headwinds and fiscal expansion as potentially raising the r-star.
• Methodological trade-offs: Surveys capture market sentiment but may overlook structural shifts (e.g., fiscal policy). Structural models explain historical drivers but lag real-time developments.

Policy Recommendations

• Hartley: Central banks should expand surveys for real-time r-star assessment.
• Carvalho et al.: Monitor fiscal sustainability and global aging trends to anticipate r-star changes.