Supply-Side Disinflation: Deregulation, Ai, And Capital Deepening
Sources: 1 • Confidence: Medium • Updated: 2026-04-02 03:48
Key takeaways
- It is uncertain whether AI will increase potential GDP more than actual GDP or vice versa because AI both drives investment demand (GPUs, data centers) and may leak demand abroad due to overseas GPU production.
- Miran places his neutral-rate estimate near the lower end of the range reflected in the SEP while still within that range.
- Oil-price spikes raise headline inflation immediately but have little inflationary consequence 12–18 months later, so central banks should generally look through oil shocks.
- Financial innovation and advances in financial technology help allocate capital more efficiently and contribute to long-run growth in human prosperity.
- Recent measured inflation is overstated by roughly 30–40 basis points because of measurement quirks in portfolio management services.
Sections
Supply-Side Disinflation: Deregulation, Ai, And Capital Deepening
- It is uncertain whether AI will increase potential GDP more than actual GDP or vice versa because AI both drives investment demand (GPUs, data centers) and may leak demand abroad due to overseas GPU production.
- AI, deregulation, and capital deepening can expand supply capacity so that faster demand growth is less inflationary.
- Supply-side policy affects inflation through two main channels: lowering prices via lower production costs and competition, and changing inflation pressure via the output gap.
- Easing regulations can raise potential GDP more than actual GDP because firms can increase utilization of existing capital (for example, running facilities longer) without requiring much new investment demand.
- The 1990s telecom-driven productivity boom plausibly pushed both demand (via investment activity) and supply (via higher productivity) such that the net output-gap effect could be roughly neutral.
- Recent tax legislation included full expensing for equipment and full expensing for R&D.
Neutral Rate (R*) Level And Drivers
- Miran places his neutral-rate estimate near the lower end of the range reflected in the SEP while still within that range.
- Miran lifted his policy-rate projection by 50 basis points in the March SEP due to intervening inflation data, placing his projected rate path around neutral.
- Miran estimates the neutral policy rate at roughly 2.5% to 2.75% and assesses current policy as about a percentage point above neutral (modestly restrictive).
- Working-age population growth swung sharply from rapid growth in 2021–2023 to near-flat growth more recently.
- Near-flat working-age population growth is expected to weigh on interest rates.
- AI-driven productivity gains raise the neutral rate, but this can be offset by demographics (slowing population growth) and an improving fiscal deficit that reduce equilibrium interest rates.
Policy Reaction Function: Lags And Oil-Shock Look-Through
- Oil-price spikes raise headline inflation immediately but have little inflationary consequence 12–18 months later, so central banks should generally look through oil shocks.
- Higher oil prices are more likely to weaken the economy 12–18 months later by diverting household spending to gasoline and thus can raise unemployment rather than generate persistent inflation.
- Monetary policy affects growth, unemployment, and inflation with long lags of roughly 12–18 months.
- Policy should move back to neutral over the course of this year because the economy does not need restraint and does not need 2021–2022-style stimulus.
Stablecoins: Regulatory Plumbing And Macro Linkage
- Financial innovation and advances in financial technology help allocate capital more efficiently and contribute to long-run growth in human prosperity.
- Stablecoins add more value for payments than for savings in open capital-market economies, but could drive large adoption in capital-control, unbanked, or high-inflation jurisdictions seeking dollar exposure.
- Introducing 'skinny master accounts' for stablecoins is under active examination and rulemaking, with issues such as ACH access and account size caps under review.
- If stablecoins grow optimistically, they could generate sizable foreign inflows into dollar-denominated savings that would weigh on the neutral rate, potentially on the order of half the magnitude of the late-1990s/early-2000s global savings glut.
Inflation Measurement And Market-Based Expectations
- Recent measured inflation is overstated by roughly 30–40 basis points because of measurement quirks in portfolio management services.
- Forward inflation expectations from inflation swaps one to three years out have been largely unaffected by recent oil developments, and many are lower than at the January FOMC meeting.
- Miran raised his current-year headline inflation projection to 2.7% because of the oil shock.
Watchlist
- It is uncertain whether AI will increase potential GDP more than actual GDP or vice versa because AI both drives investment demand (GPUs, data centers) and may leak demand abroad due to overseas GPU production.
Unknowns
- What is the empirical basis and exact methodology for the claimed 30–40 bps inflation overstatement from portfolio management services, and does it show up in PCE, CPI, or both?
- Do official labor-market indicators (unemployment duration, hiring rates, new-entrant outcomes) confirm a three-year gradual weakening trend, and how large is it?
- How stable are medium-term inflation expectations across multiple measures (swaps, breakevens, surveys) following the oil shock, and for how long must they remain stable to support the 'look-through' stance?
- Does the oil shock primarily transmit into weaker real consumption and higher unemployment over the next 12–18 months, as claimed, and what thresholds would falsify that transmission path?
- What is the content and empirical identification in the cited Fed staff paper on deregulation and inflation, and which sectors drive the estimated 0.3 pp/year effect?