Overview of Bloomberg’s Odd Lots conversation with Adam Posen
In this Jackson Hole episode, Adam Posen argues that the Fed is in a fragile spot: inflation is still too high, policy may still be too loose, and if the Fed doesn’t tighten further, things could get “messy” fast. He is somewhat more optimistic about the tone of the latest Jackson Hole speech than earlier Fed communications, but he still thinks the central bank is trying to preserve too much discretion and not enough clarity. The second half of the conversation widens out to central bank governance, political pressure on the Fed, and whether AI is really changing jobs and productivity as quickly as headlines suggest.
Main takeaways on the Fed and inflation
Posen’s read on the Jackson Hole speech
- He rates the speech roughly a B-minus, upgraded to a B-plus because it was more normal and less chaotic than prior remarks.
- His main positive takeaway: the speech made clear that core PCE at 2% is still the target and included more mainstream views on wages and financial conditions.
- His main concern: the Fed chair still seems to want maximum last-minute discretion, rather than pre-committing to a clear rule or framework.
Why he thinks the Fed may need to hike
- Posen says the speech implied a hawkish conclusion without stating it outright:
- inflation is still running too hot,
- the trend is not clearly improving,
- and policy looks insufficiently restrictive.
- He expects:
- no hike in September to be unlikely, and
- a hike by December very likely, with the possibility of 75–100 bps more tightening within six months.
- His broader inflation view: the U.S. is still likely to be stuck in the 3.5%–4.5% inflation range for a while, with upside risk.
Fed mistakes and accountability
- Posen argues the Fed, especially under Powell, did get important things wrong:
- it was late to hike in 2022,
- it cut too early in 2024,
- and it has remained behind the curve.
- He emphasizes that inflation has been above target for well over four years, and that the Fed’s repeated insistence that labor markets would collapse proved overstated.
Central bank independence and the danger of “fiscal dominance”
What counts as political interference
- Posen says some amount of public Fed criticism from presidents is normal.
- What becomes dangerous is when politicians effectively tell the Fed:
- keep rates low to help sell government debt, or
- alter the Fed’s institutional structure to make it more politically pliable.
- That’s the core of fiscal dominance: monetary policy bending to the financing needs of the state.
Why the Fed’s structure matters
- He notes that the Fed is unusual because the chair often carries a lot of informal authority, even though the system is supposed to be committee-based.
- He worries that Powell’s style has discouraged dissent:
- there may be an unwritten norm that only a few members dissent,
- and the chair “never loses.”
- In his view, that can be unhealthy because it reduces real debate and makes the committee more dependent on one person.
Communications, task forces, and “family fight”
The point of the new communications review
- The episode discusses a new Fed communications task force linked to Mervyn King and Peter Fisher.
- Posen thinks this group could produce more radical recommendations than people expect.
- His guess: they may be skeptical of the current Fed habit of publishing too much forward guidance and too many signals.
What Posen thinks the problem is
- He believes some of the current communication toolkit:
- dot plots,
- detailed guidance,
- and highly specific forecasts may now be generating noise rather than clarity.
- At the same time, he says that the phrase “family fight” sounds nice, but Powell may have used it partly to justify too little dissent.
- Posen’s view:
- healthy disagreement is good,
- but it must not turn into a system where the chair quietly controls everything.
AI, jobs, and productivity
Why economists are joining AI firms
- Posen says the migration of economists into AI companies is not unprecedented.
- The attraction is a mix of:
- big money,
- prestige,
- and the feeling of being part of a truly transformative technology.
- But he also warns that once economists join these firms, their independence becomes harder to preserve.
Are jobs being displaced already?
- His answer: not yet, at least not clearly.
- He sees more evidence of productivity improvements than labor market destruction so far.
- He mentions that even jobs often assumed to be easy targets for AI, like coding, are not yet showing the expected collapse.
“Messy jobs” and the human advantage
- Posen strongly likes the idea that many jobs are too messy and context-dependent to be automated quickly.
- Examples he mentions or implies:
- truckers do far more than drive,
- musicians or service workers adjust to human moods,
- many jobs involve judgment, relationships, and local knowledge.
- His broader point:
- AI may be useful immediately for individual tasks,
- but major labor displacement likely takes time because firms must reorganize around the technology first.
GDP measurement and the AI economy
Is AI being undercounted?
- The discussion touches on whether chip production and AI-related activity are being undermeasured in GDP.
- Posen says this may matter at the margin, but it does not change the big macro story.
- His view:
- market participants already see the money being made,
- and measurement issues are worth studying,
- but they are not yet evidence of a giant hidden AI boom.
Why the productivity boom may still be ahead
- He argues the big economic gains from AI may come later, once businesses redesign workflows around it.
- That’s similar to earlier general-purpose technologies:
- the invention shows up first,
- then the organizational transformation drives the real gains.
Bottom line
Adam Posen’s core message is that the Fed is still facing an inflation problem, and if it hesitates much longer, the outcome could become genuinely messy. He thinks Powell’s recent speech was better than earlier ones, but still too vague and too discretionary. Beyond the Fed, he sees AI as promising but not yet showing the dramatic labor-market disruption that some predict, with human judgment and messy real-world jobs still proving surprisingly resilient.