Former Fed Vice Chair: The Biggest Obstacle to Shrinking the Balance Sheet Isn't Economics, It's Politics
Bankers don't avoid the discount window because they fear the penalty rate — they fear that two years from now Congress will publish the borrowing records and their boss will be called to testify.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Burns feared the cost; Volcker accepted it
Kohn reduces the difference between the two chairs to their judgment about cost. Burns wouldn't hit hard because he feared the unemployment cost was too high and too persistent, and because he himself thought monetary policy wasn't that effective — fiscal policy, cost-of-living clauses and union contracts were what mattered. Volcker said the cost had to be paid, but it was one-time, and once paid, inflation expectations would shift. Kohn's judgment: Volcker didn't push all the way until inflation fell very low before stopping — he backed off in the second half of 1982. The official reason was that the relationship between monetary aggregates and income and prices had broken down because of deregulation, but Kohn thinks the real reason was that inflation expectations had already come down and financial stability was at risk — several big banks with Latin American loans couldn't hold up.
— Donald KohnGreenspan held the line first, and only then got productivity
Kohn deliberately pushes the timeline earlier: Greenspan's hardest stretch was actually the years after the 1991 recession. The fed funds rate had already been cut to 3%, an extreme low for the time, and was held there for a while; the pressure to ease further was enormous, and it was the first "jobless recovery." George H.W. Bush publicly demanded rate cuts and, after leaving office, told David Frost in an interview, "I reappointed him, and he disappointed me." On the Senate side, three Democratic senators — Sarbanes, Sasser and others — took turns berating him at hearings. Greenspan held firm, then in 1994 reversed and hiked, engineering a soft landing, and eased a bit more in early 1995. Kohn says it was precisely this resistance that locked in the downward pressure on inflation, which is what made the later productivity call possible.
— Donald KohnGreenspan wouldn't give long-run forecasts, but always gave a narrative
Kohn says Greenspan never submitted forecasts two or three years out, on the grounds that "I don't know what will happen two or three years from now," so Kohn and Mike Prell had to lobby him every time just to occasionally get the staff forecast into the semiannual monetary policy report. But in his monetary policy testimony Greenspan always had a narrative: what was happening in the economy, why it was happening, where the risks were, what the Committee thought and why it thought so. Kohn considers this more important than the dot plot — not a string of dots, but letting you know what the Committee is looking at. When Kohn wrote his own speeches, he imagined himself delivering the right-hand side of the Taylor rule: here is my expectation for output, employment and inflation — let the market derive the rate itself, rather than my stating it.
— Donald KohnWarsh's two worries are, to Kohn, both empirical questions
Kohn restates Warsh's two concerns about over-communication: first, that Fed communication suppresses the market's ability to interpret incoming data; second, that even if forward guidance isn't a commitment, it makes the Committee more reluctant to change its tune and therefore makes policy harder to change. Kohn's attitude: I don't know whether that's true — these are empirical questions. He cites research referenced by Bill Nelson showing that market forward guidance did not suppress the market's response to data, at least during 2010 to 2012. His conclusion isn't to scrap communication but to change its form: you don't have to say what you'll do at the next meeting or two, but you must give enough information for the market to infer the reaction function — because the market's response to data already embeds its guesses about how the Fed thinks, and if you don't say, they may guess wrong.
— Donald KohnInflation expectations aren't unanchored, but the near end is more sensitive
Kohn looks at TIPS-implied inflation expectations: the five-by-five forward has been flat as a board, and the second five years of the ten barely moved, holding a touch above 2%. But the near end is different — when oil prices rose, five-year inflation expectations clearly picked up, then fell back after that memo came out. His judgment: long-run expectations are still anchored, but short-run expectations may have become more sensitive because of the 2021-2022 episode, and that's a worry. He watches two things at once: wages and labor costs, and expectations. Labor costs are behaving well so far, and productivity from 2019 to 2026 has been roughly a bit above 2%, which he guesses has little to do with AI. He says explicitly that he was not comfortable with the December "insurance" cut, and wants confirmation that the supply shock is fading and underlying inflation is coming down.
— Donald KohnPeople hate inflation because wages are earned and price increases aren't
Kohn uses a psychological mechanism to explain why inflation is so politically lethal: if my wage goes up, that's because I'm a good worker, it's a reward for me; price increases just erode that raise. Economists will say that after inflation, the real wage has barely risen, but the worker will say that without prices eating my purchasing power, this would be a real raise. So short-run inflation expectations — the Michigan survey, the New York Fed survey — are extremely sensitive to gasoline and food, things you buy every day and can't do without. Kohn cites Greenspan's definition of price stability: households and businesses don't have to pay attention to inflation, they just go about their labor-market and product-market decisions as usual. He thinks the US has to some degree lost that state.
— Donald KohnSame Jackson Hole, two people reach opposite conclusions
Kohn identifies the real disagreement in the balance sheet task force: Jeremy Stein and Robin Greenwood argue for a large balance sheet, on the logic that the market pays a premium for liquid assets, and that if the Fed supplies more liquid assets it will discourage banks from doing maturity transformation and issuing less commercial paper, thereby stabilizing the financial system. Raghu Rajan, at Jackson Hole a few years later, reached the opposite conclusion: when the Fed supplies liquidity, it actually encourages banks to supply runnable liquidity to the private sector. Kohn says he doesn't know who's right, but his reaction to Rajan's paper was: two things were happening at once in that period, QE and zero rates, and a lot of the liquidity supply actually came from zero rates and the expectation that they would stay low for a long time — Silicon Valley Bank issued uninsured demand deposits to buy long-dated MBS and took on enormous interest rate risk. He thinks the real failure was that supervisors didn't warn about interest rate risk.
— Donald KohnHalf the discount window stigma comes from Congress
Kohn says the discount window stigma has two sources. One is market perception: such-and-such a bank has to go borrow — is something wrong? The other is a political source few people talk about: borrowing from the Fed in 2008 and 2009 was treated as a bailout, and in 2010 Dodd-Frank required that borrowing information be published two years later. He took part in a roundtable Bill Nelson organized, asking bank treasurers why they didn't use the window, and the answer was: "My boss said if he gets called to testify before Congress over this, I'm fired." Kohn says the Fed can handle the market half of the stigma; the political half is harder — you have to convince politicians that the discount window isn't a bailout and that it even carries a somewhat punitive rate. He was grilled by House and Senate committees in 2009 and 2010 over the Fed's lending, so he's especially sensitive to this.
— Donald KohnIn their own words · checked verbatim
And I think another lesson was one reason that the Fed under Arthur Burns didn't take the steps necessary was they were afraid of the cost. They were afraid that the unemployment costs would be too high, too persistent.
Donald Kohn4:12
After he left office, George Bush, in his interview with David Frost, retrospective, said of Alan Greenspan, I reappointed him, he disappointed me.
Donald Kohn16:38
But I think the really important point is that Greenspan always had a narrative about what was going on and an explanation in his monetary policy testimonies.
Donald Kohn23:48
I think part of this phenomenon is when I get a wage increase, that's because I'm a great worker. I'm being rewarded for my great stuff. Price increases just erode that wage increase.
Donald Kohn41:24
When households and businesses don't have to pay attention and don't pay attention to inflation, they just go about making their decisions about the labor market, about the product market, without worrying about this inflation going up.
Donald Kohn42:28
And they were asked, why don't you use the window? And a lot of it is my boss tells me that if he's called before Congress and has to testify, I'm fired.
Donald Kohn1:00:40
So I think I don't buy the footprint argument particularly.
Donald Kohn1:05:57
Figures
| One-day drop in the 1987 crash | More than 20% (Kohn says about 22%) | 12:33 |
| Fed funds low after the 1991 recession | 3% | 15:36 |
| Second five years of ten-year TIPS-implied inflation expectations | A touch above 2% | 36:19 |
| Productivity growth, 2019 to 2026 | About 2% or a bit above 2% | 37:21 |
| Year Kohn testified on interest on reserves | June 2004 | 43:32 |
| Legislation authorizing interest on reserves | The 2006 Financial Services Regulatory Relief Act | 44:34 |
| Original effective date for interest on reserves | 2011, implemented early in 2008 | 44:34 |
| Dodd-Frank disclosure requirement for Fed borrowing information | Published two years later | 1:00:40 |
Glossary
- interest on reserves
- Interest the Fed pays banks on reserves held at the central bank; the core tool for rate control after 2008.
- asymmetric corridor system
- The Fed's pre-2008 rate corridor, with a movable ceiling and a floor locked at 0%.
- ample reserves
- The Fed's current operating framework, in which reserve supply far exceeds demand and the rate is driven by the interest rate on reserves.
- discount window
- The Fed's standing facility for lending to depository institutions, long underused because of stigma.
- five by five
- The five-year inflation expectation starting five years out, used to strip out near-term noise and see long-run expectations.
- market-based core
- A core inflation measure that strips out prices relying on estimation, such as financial services; Kohn prefers it.
How to listen
Investors doing macro trading, rates and dollar-asset allocation, and practitioners who care about the Fed's operating framework and communication machinery.
The opening 0:00-3:05 small talk about old times and the Purple Heart anecdote can be skipped.