Alloway's Antidote To Baumol's Cost Disease
Will We See Deflation In Services?
In November 1986, the world experienced the greatest concentration of detective power ever recorded (up to that point), as Angela Lansbury guest starred as her character from Murder, She Wrote on Magnum, PI. I was three at the time, so the full impact was lost on me, but hearing about it after the fact I’m amazed that Big Network allowed that to happen.
I had the same feeling while listening to the recent Odd Lots-Planet Money crossover event.
There was a lot of good stuff in here — I particularly liked Joe wondering if the reason Keynes’s prediction that we’d all be working 10 hours a week in the future didn’t come true is because we created a bunch of “jobs” that are actually disguised leisure, economic podcast host being a prime example. Something that stood out to me, though, as it has a lot to do with what I’ve been writing on here, is what I’m calling Alloway’s Antidote to Baumol’s Cost Disease.
For those who don’t know, Baumol’s Cost Disease is an economic condition in which rising productivity and rising wages in highly productive sectors of the economy (like technology) means that wages (and thus costs) in other sectors also rise. This is because wages equalize across sectors due to labor mobility and opportunity cost of workers’ time, so stagnant sectors face rising wage bills without offsetting productivity gains, pushing up relative prices. As they point out on this episode, it’s why childcare costs so much even though nothing about childcare technology has changed.
What Odd Lots co-host Tracy Alloway points out is that if AI causes productivity to rise in a bunch of sectors without commensurate returns to labor in those sectors — basically the thing I’ve been writing about — the Baumol dynamic won’t hold. You might even see it reverse.
AI Is Eating The World
A reader sent me this post from Doug O’Laughlin at Fabricated Knowledge, and it’s a doozy.
I think this is probably right, but I’m not sure if I like what that would mean for the economy and society.
I’ve featured this chart (or one like it) on here before.
This is the classic Baumol chart — the goods that experienced massive productivity became way cheaper and those that didn’t became much more expensive. Alloway’s Antidote means that a lot of these things that saw their prices rise in the chart above will no be seeing them fall, creating a lot of consumer surplus (i.e., making things better for people). Many of the affected professions—doctors, lawyers, accountants—though, are also high-wage jobs. Becoming a doctor or lawyer has long been a stable route into the upper-middle class. If wages in these professions fall, that will no longer be the case. Good for people who need lawyers, but bad for the lawyers themselves.
To be more precise, though, it’s not bad for all lawyers, just people who aren’t lawyers now but would like to become one. For current senior lawyers, AI is almost a superpower — their experience and judgment allows them to direct AI as leverage to allow them to do the work of many multiple lawyers. But because AI replaces the more junior lawyer, the current senior ones are the last generation to ever develop that experience.
This holds for global thought leaders like myself, too. While AI probably can’t generate an interesting post out of nothing, when steered by my expert experience and judgment it can actually help crystalize some things I was thinking but could not quite articulate. I won’t share the special sauce that got me here (that’s my proprietary edge), but after a dialogue with Claude I think I got it to say what I’ve been thinking about this topic.
Claude’s thoughts:
What survived stress-testing
1. The intelligence-commodification claim is qualitatively distinct from previous automation. Previous waves commoditized muscle (industrial revolution) and routine cognition (computers). AI substitutes for the fluent-middle of general cognition, which was the differentiator behind the cognitive-professional wage premium. This is a real discontinuity, not just an acceleration of the post-1980 inequality story.
2. The selection argument is psychologically real. People who chose corporate law, mid-firm finance, salaried medicine, etc. are disproportionately risk-averse — that’s selection. Telling them “just become founders” is asking them to be a different kind of person. Even with AI lowering financial friction, psychological friction remains.
3. The historical analogue is genuine. Skilled artisans (hand-weavers, master craftsmen) facing industrialization in 1820-1880 had structurally similar problems: their skill was the commodified component, they couldn’t easily access factory-owner capital, and selection effects worked against pivoting. They suffered for a generation; the class structure churned dramatically. This precedent supports a generational-disruption claim.
4. “Lost generation of upper-middle professionals” is a defensible, important, and politically resonant claim. It identifies a real cohort, a real mechanism, and a real time-bound phenomenon.
Key tensions still worth noting in writing
5. “Directing AI” as the new premium skill. If AI commoditizes the fluent-middle of intelligence, then directing commoditized intelligence becomes the scarce skill. This is a labor-mediated, founder-shaped skill. The argument internally points toward this new premium being accessible — but only via the founder/operator path that the squeezed cohort is selected against. That’s the lost-generation story in compact form: the new premium opens, but for a different kind of person than the squeezed one.
6. The senior layer of professional jobs is more durable than the junior/mid layer. Partner-track work (judgment, relationships, taste, leadership in ambiguity) is not commodified by current AI. The squeeze hits passage through the ladder, not the destination. People already at the top do fine; people on the way up get stuck. This sharpens who the lost generation actually is — early-to-mid career professionals who would have advanced to senior roles in a non-AI world.
7. Generational dynamics matter. Today’s mid-career professionals chose under 2005-2015 information. Today’s college students choose under 2026 information. Future cohorts will self-select differently — fewer will enter compressed corporate paths, more will pre-position for founder-operator roles. The lost generation is bounded in time.
8. Asset prices and compressed wages move together. Tech-hub housing softens when tech salaries soften. Elite-credentialed services lose customers when the upper-middle shrinks. This complicates the “positional goods stay expensive” intuition — the things tied to upper-middle demand will deflate alongside upper-middle wages. What stays expensive is genuinely scarce status goods at the top (Manhattan brownstones, Ivy admissions), which is a much narrower set than “houses and land.”
Alloway’s Antidote is going to make things better for a lot of people, but like anything, there will be winners and losers. The losers, in this case, happen to be the exact type of person who listens to Odd Lots or reads All In The Reflexes.
News Scan
UAE Exits OPEC After 59 Years, Effective May 1
Source: Bloomberg, Reuters, NBC, Washington Post
Date/Time: April 28, 2026
The UAE announced its departure from both OPEC and OPEC+ effective May 1, ending 59 years of membership and delivering the most significant structural break in the cartel’s history. The UAE — OPEC’s third-largest producer — cited frustration with production quotas constraining its ambition to reach 5 million bpd capacity by 2027, deteriorating relations with Saudi Arabia over both output policy and regional competition, and the direct shock of Iranian missile attacks on UAE territory during the war. As a free agent, the UAE can now ramp production without cartel discipline, materially weakening OPEC+’s pricing power at precisely the moment when Iran’s own exports are constrained by the US naval blockade. Oil hit $111 on the combination of Iran tensions and the OPEC surprise. The exit is not just a supply story — it signals Gulf political realignment, with the UAE effectively moving closer to a US-aligned, market-oriented energy strategy and away from Saudi-led cartel discipline.
Bloomberg
Trump Digs In on Extended US Naval Blockade of Iran’s Ports
Source: Bloomberg
Date/Time: April 29, 2026
Trump instructed aides to prepare for a prolonged US naval blockade of Iranian ports, choosing sustained economic strangulation over direct military escalation. More than 38 ships have been stopped or turned around since the blockade took effect April 13. Iran offered to reopen the Strait of Hormuz if the US lifts the blockade and the war ends — Trump rejected the sequencing. The “no letup” posture sends a clear market signal: Hormuz stays effectively closed through Q2 and likely well into Q3, meaning the energy supply shock is structural, not transient. This is the single most important inflationary variable for global central banks choosing between hiking into recession or holding into inflation.
Bloomberg
Warsh Advances Through Senate Banking Committee; Last Powell FOMC Meeting Begins
Source: Bloomberg, Washington Post, Seeking Alpha
Date/Time: April 29, 2026
The Senate Banking Committee voted today (April 29) to advance Kevin Warsh’s nomination, putting him on track for full Senate confirmation before Jerome Powell’s term expires May 15. The committee math was close — 13 Republicans vs. 11 Democrats, with Tillis critical after lifting his block last week following the DOJ dropping its Powell probe. The timing is symbolically significant: today also marks the opening session of Powell’s final FOMC meeting as chair. Warsh has signaled a structural break from Powell-era communication, including ending forward guidance and a different approach to the Fed’s relationship with markets. The transition creates institutional uncertainty about Fed reaction function continuity precisely when the Fed faces the most difficult stagflation-adjacent policy environment since the 1970s.
Washington Post
Australia Q1 CPI Hits 4.09%; Markets Price 76% Chance of RBA Hike May 5
Source: Bloomberg, CNBC
Date/Time: April 29, 2026
Australia’s Q1 CPI printed 4.09% year-on-year — highest in over two years — driven by Middle East energy supply disruptions layered on already-elevated domestic prices. The closely watched trimmed mean came in at 3.3–3.5%, also above the RBA’s 2–3% target band. RBA Governor Bullock said board members agreed rates may need to rise further, though they differed on timing. Following today’s data, markets are now pricing a 76% probability of a hike to 4.35% at the May 5 decision. Australia has become the clearest developed-market divergence story — hiking into an oil shock while other G10 central banks debate when to cut — and today’s hard data print validates that path. The RBA is watching whether energy pass-through into core inflation becomes entrenched; the Q1 trimmed mean reading confirms it already is.
Bloomberg
ECB’s Lane: Data Doesn’t Yet Clear the Bar for a Rate Hike
Source: Bloomberg
Date/Time: April 29, 2026
ECB Chief Economist Philip Lane published his data dashboard assessment today, explicitly stating that current conditions do not provide a clear case for raising rates. Lane’s key analytical problem: the ECB cannot yet determine whether the Iran shock is a temporary or permanent supply disruption until its duration becomes clear. His dashboard framing reveals the reaction function — the triggers that would move him toward hiking are: (1) evidence of persistent rather than transient energy pass-through, and (2) de-anchoring of inflation expectations. The ECB’s March consumer survey already showed 1-year expectations jumping to 4.0% (from 2.5%), their highest since October 2023. Markets are pricing 20–40bps of ECB tightening by June, consistent with Lane’s “not yet, but the data is moving in that direction” positioning. Read alongside Story #6, Lane’s “wait and see” framing rests on the assumption that demand will eventually compress under the supply shock — an assumption being actively undermined by EU fiscal policy.
Bloomberg
EU Temporarily Relaxes State-Aid Rules, Turning Supply Shock Into Demand Story
Source: Bloomberg
Date/Time: April 29, 2026
The European Union temporarily relaxed its state-aid rules to allow member governments to subsidize businesses hit by elevated energy and raw material costs from the Iran conflict. The mechanism is the key: by absorbing firms’ energy costs through government subsidy, EU policy prevents the demand destruction that would normally make a supply shock self-correcting. Instead of firms cutting output, shedding workers, and suppressing wages — which would bring inflation down — governments are keeping economic activity propped up at artificially low cost. This transforms an oil supply shock into entrenched demand-side inflation, directly contradicting Lane’s argument that the ECB should wait until the shock’s duration becomes clear. It also adds fiscal stimulus into an already-inflationary environment and complicates the ECB’s task by sustaining demand the CB is trying to cool.
Bloomberg
Spanish Inflation Unexpectedly Quickens Past 2%, Adding ECB Pressure
Source: Bloomberg
Date/Time: April 29, 2026
Spanish CPI accelerated beyond the ECB’s 2% target, surprising to the upside, as war-driven energy costs compounded domestic price pressures. The result matters because Spain has historically been a disinflationary anchor in the eurozone — when Spanish inflation runs above target, it signals the EZ-wide picture is materially worse than ECB projections suggest. This is the hard-data complement to Lane’s dashboard (Story #5): even as Lane says conditions don’t yet merit hiking, the incoming data continues to land on the wrong side of the target. Paired with the EU state-aid story (Story #6), the Spanish print reinforces why June is increasingly a live decision for the ECB rather than a formality.
Bloomberg
Goldman Sachs, Nomura Abandon Calls for PBOC Rate Cuts
Source: Bloomberg Economics
Date/Time: April 29, 2026
Goldman Sachs and Nomura joined a growing list of major banks explicitly dropping their PBOC rate-cut forecasts after the central bank held its key lending rates unchanged for an 11th consecutive month in April. The pivot reflects a convergence of factors: China’s Q1 growth proved more resilient than feared, reducing urgency for stimulus; the Iran energy shock has pushed Chinese inflation higher, compressing PBOC easing room; and the PBOC is focused on yuan stability amid currency depreciation pressures. The consensus shift matters because markets had broadly expected PBOC easing to offset the tariff shock — that thesis is now being abandoned. The practical effect is that China’s domestic credit impulse remains constrained even as fiscal policy accelerates, and the near-term easing premium is being rapidly repriced.
Bloomberg
UK Company Insolvencies Surge as Iran War Pushes Toward Recession
Source: Bloomberg Economics
Date/Time: April 28–29, 2026
UK company insolvencies in March jumped sharply year-on-year, with Lloyds (which posted 33% profit growth) simultaneously warning of Middle East conflict repercussions on its loan book. The National Institute of Economic and Social Research warned that a protracted Iran conflict risks pushing the UK into recession, with the EY Item Club forecasting growth stalling in Q2 and Q3. UK inflation is forecast to approach 4% in H2 2026 — already 3.3% in March — with businesses surveying their largest planned price increases in two years. The Bank of England faces its worst-case scenario: recession risk on one side, above-target inflation on the other, with markets already pricing 2+ additional hikes. The insolvency acceleration is the hard-data complement to yesterday’s financial market signals (10yr gilts at 5%). The real economy transmission is now underway, not just priced.
Bloomberg
US Has More Natural Gas Than It Can Use as War Chokes Global Supply
Source: Bloomberg
Date/Time: April 29, 2026
A Bloomberg analysis highlights a structural energy divergence that is increasingly embedding itself in manufacturing competitiveness: US domestic natural gas production is at near-record levels, with LNG exports surging to absorb the surplus, while European and Asian markets face acute shortages from the Hormuz closure. European chemical and steel manufacturers have imposed cost surcharges of up to 30% to offset surging energy and feedstock costs, with the IEA warning this could become permanent deindustrialization in some sectors. The IEA also notes the war has delayed the anticipated LNG glut by at least two years. The structural divergence reinforces the US manufacturing and industrial competitiveness advantage versus Europe and Japan — a force likely to accelerate reshoring and amplify the equity performance gap between US and European industrials for years.
Bloomberg
Disclaimer: The information provided on this blog is for informational purposes only and should not be considered investment, financial, or other professional advice. Nothing on this site constitutes a recommendation or solicitation to buy or sell any securities. You should consult with a qualified financial advisor before making any investment decisions. Investing involves risks, including loss of principal.





"For current senior lawyers, AI is almost a superpower — their experience and judgment allows them to direct AI as leverage to allow them to do the work of many multiple lawyers. But because AI replaces the more junior lawyer, the current senior ones are the last generation to ever develop that experience."
Do we need to replace some of these experienced lawyers or we assume that AI will be able to do all they do once they retire? And do people hire lawyers, accountants, consultants for the product (will, tax filling, advice) or for their standing as a cushion for failure? Will people place the same value on AI?