Long time reader, probably first ever comment. This is great, just the right level for me (non economist). I was going to say, please do more of these from time to time, but with a paid up GPT5.1 pro subscription, I can ask my assistant to turn your essays that I often struggle to fully understand into "noddy level".
"...the purchasing power of money is inversely proportional to the price of goods and services."
Exactly. Because every transaction has to have both a buyer and a seller. I.e., if the price of goods and services is rising in money terms, by logical necessity, the price of money is declining in terms of goods and services. Which is why interest rates are NOT 'the price of money.'
That analysis is straight out of Milton Friedman (and his son David's 'Price Theory' textbook).
Let's imagine a central bank that uses the price of money as their policy instrument. (The actual price, not the interest rate.) They offer to buy or sell essentially unlimited amounts of their own money in exchange for some assets in order to drive the price of their own money to their desired point. (Singapore's MAS actually does this!)
So if MAS implements a policy of eg 1% per year appreciation for the SGD compared to the USD (and the market anticipates the execution of that policy), then in a frictionless-spherical-cow model of the economy, Singapore's interest rates will be roughly 1% over prevailing US interest rates. Any deviation, and arbitrageurs can make free money.
In practice, you have a few frictions, but it's relatively close to true.
Pagerank is the famous algorithm that drove the early success of Google. AI pagerank is the next scarier iteration of the same algorithm as perfected by OpenAI and some other firms.
I input your first prompt into Gemini Flash, and the first two references were you and Doug Irwin. Roughly similar answers, but Gemini also mentions gold sterilization and private hoarding.
Also, the second prompt references The Midas Paradox as the "definitive 'gold-market approach' to the Great Depression."
Umm, I think you missed the point, which is that ChatGPT is not reliable. God help us if people are using ChatGPT as a counterargument to my posts. FWIW, when I did my study back in 2008 the various free market rankings (Heritage/Fraser, etc.) agreed with me, ranking Denmark number one in the world in economic freedom, if you remove the tax/spending category and look at the other 80% of the index (privatization, trade, rule of law, deregulation, etc.)
Heritage now has Denmark #7 in Economic Freedom, far ahead of the US even if you include size of government, where Denmark scores low due to Heritage's small government bias. In areas like business freedom (#1) and property rights (#2) investment freedom and financial freedom Denmark ranks at or near the top of the world:
Obviously these are partly subjective, but note that these high rankings for Denmark are coming not from Bernie Sanders, they are coming from right wing think tanks.
Of course ChatGPT knows nothing about all of this, it's silly to rely on its views.
I asked Chatgpt the limitations of your viewpoint (just because it might be interesting to you). Answer is here:
3) Important limitations and criticisms
1. It’s not the whole story.
• Banking system collapse, runs, and the resulting destruction of the money multiplier were huge. Those banking dynamics are not just a passive consequence of deflation — they were active drivers of money supply decline. Sumner’s framework can incorporate this (as an amplifier), but it doesn’t by itself explain why banks failed so badly.
2. Direction of causation and multiple interacting forces.
• Was the primary shock a rise in money/gold demand, or was it a collapse in confidence / investment demand that reduced NGDP and raised money demand secondarily? Different interpretations imply different policy priorities.
3. Heterogeneity across countries.
• If gold-demand shifts were the single cause, we’d expect similar patterns everywhere on the gold standard. In fact the depth/duration varied a lot and that variation maps strongly to domestic policy responses (bank rescue policy, fiscal policy, when/how quickly countries left gold). That suggests institutional and policy differences mattered a great deal.
4. Role of real shocks.
• There may have been important real shocks (e.g., productivity reversals in some sectors, sectoral overinvestment from the 1920s) and structural problems (labor market mismatch) that monetary policy alone couldn’t fix. Monetary accommodation helps, but it doesn’t produce investment that’s uneconomic forever.
5. Sticky wages/prices assumption.
• The implied real effects rely on downward nominal rigidity. How sticky were wages and prices across sectors and countries? The degree of stickiness changes the quantitative implications.
6. Debt-deflation and balance-sheet channels.
• Irving Fisher’s debt-deflation theory emphasizes that deflation raises the real burden of nominal debt and bankruptcies, further depressing demand. Sumner’s story is compatible with Fisher’s amplification channel, but debt dynamics are an additional mechanism to model explicitly.
7. Empirical attribution is hard.
- Disentangling how much of the downturn was due to gold-demand driven deflation vs. banking runs vs. collapsing aggregate demand vs. fiscal tightening vs. external shocks requires careful empirical work. Many of these things happened at once.
1st point: "Sumner’s framework can incorporate this (as an amplifier), but it doesn’t by itself explain why banks failed so badly."
Yes, it clearly does. I explain how tight money caused falling NGDP which led to bank failures, as it always does.
The second criticism is apples and oranges. I am explaining why NGDP fell using a gold supply and demand framework. That will obviously causes bearish expectations, falling investment, etc. That's not an "alternative view", it's an implication.
The third point is simply incorrect. I do point out that the varying recoveries depend on differences in domestic policy---especially the timing of leaving the gold standard.
The 4th point is the Austrian argument, which is rightly rejected by most economists. Overinvestment doesn't lead to mass unemployment, as we saw in 2007. It leads to re-allocation from one sector to another. That did not happen in the 1930s, as all sectors fell. The problem was falling NGDP. The same thing happened in late 2008.
5th point is hopeless vague. I pointed out that Hoover and FDR made them stickier than in the early 1920s.
6th Again, I clearly discussed how the deflation led to debt defaults, which caused bank failures and higher gold demand.
7th. Again, All of those points are factors that might influence gold demand. I discussed the factors that I thought were the most important.
Many people overestimate the ability of AI to actually THINK. It's good at rehashing other people's arguments--like a hard working B student--but it cannot actually engage in deep thinking about macro questions. It's not there yet.
• Qualitatively convincing: It captures an essential truth: the gold standard constrained policy and monetary forces mattered a lot. Monetary contraction (and policy failure to stabilize nominal spending) is central to most modern accounts.
• Quantitatively uncertain: Whether the initial trigger was mainly a shift in gold demand (in the narrow sense) versus banking system collapse, or a collapse in confidence that raised money demand, is debated. Sumner gives a coherent mechanism that complements the Friedman-Schwartz monetary history and Fisher’s debt-deflation story, but it should be read as one piece of a multi-factor explanation.
• Policy takeaway (robust): Don’t let nominal spending collapse. That’s a robust implication — whether the trigger is gold demand, banking runs, or collapsing confidence, stabilizing nominal GDP or satisfying money demand avoids the worst outcomes.
(end)
I think some, or a lot of this, is classic Chatgpt trying to be very neutral on a debated issue. Still interesting though!
OT but interesting. The Swiss National Bank, that is Switzerland's central bank, has a balance sheet equal to 100% of GDP. They are on the cusp of deflation.
Same thing happened to the Bank of Japan, pre-COVID.
Yi Wen, 2013. "Evaluating unconventional monetary policies -why aren’t they more effective?," Working Papers 2013-028, Federal Reserve Bank of St. Louis.
To take this more serious than is probably warranted:
(1) If your money printing isn't causing (enough) inflation, you need to print more money. You can buy arbitrary assets with the newly printed money. Repeat until you reach the desired level of inflation. (Use market forecasts to guide you. Don't use a rearview mirror of measured inflation.)
(2) That has nothing to do with inflating away government debt. The goal in (1) is to get back up to the targeted path of eg 2% inflation, or for the Swiss to hit their targeted exchange rate with the Euro. Nothing beyond that.
Yes, QE, or LSAP, but also money-financed fiscal programs can be used to fight deflation or sluggish growth.
Plain ol' deficit spending, apparently backed into the US cake for the foreseeable future, can somewhat work (with monetary accommodation) but ramps up taxpayer indebtedness.
The US can avoid over-leveraging by buying back debt, through the Fed.
David Beckworth advocates QE plus tax cuts on ordinary people (QE for Main Street).
My idea is Social Security tax holidays while the Fed buys Treasuries and puts them into the SS fund. This might be a good idea anyway.
Inflation in the 2% to 3% range is OK, even maybe 3% as it helps deleverage the nation.
I don't see the moral or practical superiority of only expanding the money supply through regulated commercial bank lending.
Also, in a world of globalized capital markets...what does a lone
central bank tightening or loosening actually do?
(Maybe in a perfect world, we can balance the US budget. It happened back with Clinton. But realistically, we must devise monetary policy given chronic federal deficits and globalized capital markets, and what is actually in US interests.
If global investors accept US cash for Treasuries, and that has little or no effect on inflation, then fine, let them hold US dollars. Keep US dollars in offshore bank accounts or briefcases. Who cares?)
If the Bank of the United States had not defaulted upon its deposits, the national banking run would not have happened:
“There is nothing in the economic situation as it stood in, say, September or October 1930 that made the continued and drastic decline of the following years inevitable or even highly probable. The character of the contraction changed drastically in November 1930, when a series of bank failures led to widespread runs on banks, which its to say attempts by depositors to convert deposits into currency. The contagion spread from one part of the country to another and reached a climax with the failure on December 11, 1930 of the Bank of the United States. This failure was critical not only because the bank was one of the largest in the country, with over $200 million in deposits, but also because , though an ordinary commercial bank, it’s name had led many at home and even more abroad to regard it as somehow an official bank. Prior to October 1930 there had been no sign of a liquidity crisis or any loss of confidence in banks. From this time on, the economy was plagued by recurrent liquidity crises (which) were important because of their effect on the money supply. The Reserve System failed this test miserably. It did little or nothing to provide the banking system with liquidity, apparently regarding the bank closings as calling for no special action. The system had ample power to provide the banks with the cash their depositors were demanding. Had this been done, the bank closings would have been cut short and the monetary debacle avoided. ” (Friedman, Capitalism and Freedom, 1962)
"The Federal Reserve had the power at least to ameliorate the problems of the banks. For example, the Fed could have been more aggressive in lending cash to banks (taking their loans and other investments as collateral), or it could have simply put more cash in circulation. Either action would have made it easier for banks to obtain the cash necessary to pay off depositors, which might have stopped bank runs before they resulted in bank closings and failures. Indeed, a central element of the Federal Reserve's original mission had been to provide just this type of assistance to the banking system. The Fed's failure to fulfill its mission was, again, largely the result of the economic theories held by the Federal Reserve leadership. Many Fed officials appeared to subscribe to the infamous "liquidationist" thesis of Treasury Secretary Andrew Mellon, who argued that weeding out "weak" banks was a harsh but necessary prerequisite to the recovery of the banking system. Moreover, most of the failing banks were relatively small and not members of the Federal Reserve System, making their fate of less interest to the policymakers. In the end, Fed officials decided not to intervene in the banking crisis, contributing once again to the precipitous fall in the money supply." (Bernanke, H. Parker Willis Lecture in Economic Policy, Washington and Lee University, Lexington, Virginia, March 2, 2004, "Money, Gold, and the Great Depression".)
Yes, it is a tax. While the reduction of money hoarding is a valid objective, the cost of high inflation still exceeds the benefit. But it does make the optimal rate of inflation a bit higher than what you'd get in simple models that ignored tax evasion in the underground economy.
I'm thrilled to hear you are attracting many new readers. I hope you will take that opportunity to tout (and they will read) your online book Alternative Approaches to Monetary Policy!
Haven't read the paper, but here are a few comments on the abstract:
Conventional monetary policy is ineffective at the zero bound only if you define policy as interest rate targeting. But I don't find that to be a useful definition of monetary policy.
FDR did not do much fiscal stimulus; the recovery was largely caused by monetary stimulus.
Price fixing policies such as the NIRA were contractionary, and slowed the recovery.
This seems to suggest that if nothing is done, wages and prices will eventually correct to the new equilibrium, and no intervention is required at all. But in this context, that means deflation, and deflation incentivizes hoarding currency, which produces a feedback loop of reduced purchasing -> reduced demand for goods -> prices fall further. This has to bottom out eventually, of course, but in the mean time a lot of benefits from exchange are lost.
"This seems to suggest that if nothing is done, wages and prices will eventually correct to the new equilibrium, and no intervention is required at all."
It depends what you mean by required. Even if wages and prices are flexible in the long run, you can have a lot of suffering in the short run. So you'd still prefer a monetary policy that produced stable NGDP growth.
This response along with your other response about inflation being a tax, leads me to a question: If I remember correctly from past posts on econlog, you favor a NGDP growth target of about 4%? If RGDP growth is 2% long run, why not a 2% NGDP target, netting to 0% inflation in the long run? Is it just to target tax evasion in the underground market? Does that really matter? And would it really need to be another 2% (assuming I remember correctly that you would target 4%) or would something like a 2.5% NGDP growth target be enough, while also keeping long term inflation low?
Mainly because nominal wages are inflexible in a downward direction due to money illusion. Workers don't like negative nominal wage changes. In addition, with 4% NGDP growth (level targeting) you don't have to worry about the zero lower bound for interest rates.
I am not sure I understand the situation the first would matter (the latter makes sense). What is the circumstance in which there would be the need for negative nominal wage changes if NGDP was growing at 2%?
If RGDPg is negative (or lower than 2%), there will be some inflation to get NGDPg to 2%, so there would be no need for negative wages. If the economy is running hot, with RGDPg greater than 2% then the deflation would be in sectors other than labor. At the worst, holding labor costs constant would cool the market off, bringing it back down to the 2%.
I am probably missing something here, I realize that, it just isn't clear to me what it is.
Think of NGDP as driving the average wage change. Actual wage changes in various sectors look something like a bell-shaped curve, as each sector faces different supply and demand shocks. Unfortunately, the bell shaped curve for wage changes is truncated at zero percent, so sectors that need to see their relative wage fall have trouble making adjustments when average wage growth is quite slow.
You would be correct if all sectors had the same equilibrium nominal wage increase.
That makes sense, obviously it would still happen at 4%, but the tail would be much smaller. What is the standard deviation of this bell curve, is the tail past 2% really large enough to be problematic (if SD is less like .5%, I wouldn't be concerned with the >4SD industries; if it more like 1.5%, I see the problem)?
However, it does seem like some industries have figured this out on their own. At the last few places I have worked, 10% of my base salary is an expected bonus. The actual number can vary in both directions depending on company metrics. Like I am expecting a bonus of about 9% for 2025 (I will receive it in March). In past years it has been as high as 12%. This gives them the ability to give standard raises every year and still have negative wage growth in down years (If I get a 3% raise but only get a 5% bonus, I am making less). I realize this isn't possible in every industry, but its a free market solution to that issue.
I’m interested in that claim that the central banks held most of the gold that had ever been mined. I’ve heard in various contexts that households and temples in India actually contain a vast majority of the gold that has been ever been mined. Do you know if there’s any actually good measurement of this? (It’s definitely hard to know what is actually in the vaults below the temples in India.)
I don't recall the source, but we had good data on total central bank holdings, so it's a question of how much gold has been mined over time. Mining output rose a lot over time, so I suspect that a good portion of the mined gold was produced between 1800 and 1930, for which we'd have reasonable data. But that's just a guess.
A great economic history lesson, thank you! And I do hope the Supine (sic) Court rescues Trump from his catastrophic tariff policy!
"Supine" I fear you are correct. Even if they shoot down some of the tariffs, it will be largely symbolic.
Long time reader, probably first ever comment. This is great, just the right level for me (non economist). I was going to say, please do more of these from time to time, but with a paid up GPT5.1 pro subscription, I can ask my assistant to turn your essays that I often struggle to fully understand into "noddy level".
"...the purchasing power of money is inversely proportional to the price of goods and services."
Exactly. Because every transaction has to have both a buyer and a seller. I.e., if the price of goods and services is rising in money terms, by logical necessity, the price of money is declining in terms of goods and services. Which is why interest rates are NOT 'the price of money.'
That analysis is straight out of Milton Friedman (and his son David's 'Price Theory' textbook).
Agreed. Going on a tangent:
Have a look at Interest Rate Parity (https://en.wikipedia.org/wiki/Interest_rate_parity).
Let's imagine a central bank that uses the price of money as their policy instrument. (The actual price, not the interest rate.) They offer to buy or sell essentially unlimited amounts of their own money in exchange for some assets in order to drive the price of their own money to their desired point. (Singapore's MAS actually does this!)
So if MAS implements a policy of eg 1% per year appreciation for the SGD compared to the USD (and the market anticipates the execution of that policy), then in a frictionless-spherical-cow model of the economy, Singapore's interest rates will be roughly 1% over prevailing US interest rates. Any deviation, and arbitrageurs can make free money.
In practice, you have a few frictions, but it's relatively close to true.
As Josh Hendrickson wrote: "I think one could argue that it is Scott who has carried on the Chicago tradition in monetary economics more than any other monetary economist." https://www.economicforces.xyz/p/im-teaching-a-course-on-bitcoin-why
Scott, on one hand, your AI pagerank is too low. On the other hand, you are standing on the shoulders of the giants of monetarism.
I am certainly standing on Friedman's shoulders, but what is AI pagerank?
Pagerank is the famous algorithm that drove the early success of Google. AI pagerank is the next scarier iteration of the same algorithm as perfected by OpenAI and some other firms.
LLM AIs are structured for helpfulness not truthfulness. That is a big problem.
https://hollymathnerd.substack.com/p/the-golden-retriever-of-software
I input your first prompt into Gemini Flash, and the first two references were you and Doug Irwin. Roughly similar answers, but Gemini also mentions gold sterilization and private hoarding.
Also, the second prompt references The Midas Paradox as the "definitive 'gold-market approach' to the Great Depression."
That's good to hear. I've relied on ChatGPT out of laziness, but I probably need to explore other AIs.
Speaking of ChatGPT. Please read the final sentence.
Denmark is best described as a social democratic / universalist welfare state with market coordination, often summarized as “flexicurity.”
Key features:
Very large welfare state
High public spending (~50% of GDP):
Universal healthcare, education, childcare, pensions
Very high taxes:
Among the highest tax burdens in the world
Broad-based, progressive, explicitly funding redistribution
Strong labor institutions:
High union density
Sector-wide collective bargaining
Strong worker protections outside of employment (income support, retraining)
Active state, not a minimal one:
Heavy regulation
Extensive public services
State actively shapes markets rather than retreating from them
Where the “neoliberal” confusion comes from
People point to a few real things and overextend them:
Ease of doing business
Open trade
Few formal employment protection laws (easy to fire)
But:
Job insecurity is offset by very generous unemployment benefits
The state absorbs risk that markets create
Labor power is preserved institutionally, not via strict job-lock laws
This is not neoliberalism; it’s risk socialization + market dynamism.
A more accurate framing
Denmark is:
Market-friendly
Capitalist
Highly redistributive
Institutionally anti-neoliberal in outcomes
If you forced a label, Denmark is closer to:
Social democracy
Coordinated market economy
Embedded liberalism (postwar model, not neoliberalism)
Bottom line
Calling Denmark “the most neoliberal country”:
❌ Misrepresents neoliberalism
❌ Ignores welfare state size and labor power
❌ Confuses market efficiency with market dominance
It’s more accurate to say Denmark shows how far you can go against neoliberalism while still using markets effectively.
Umm, I think you missed the point, which is that ChatGPT is not reliable. God help us if people are using ChatGPT as a counterargument to my posts. FWIW, when I did my study back in 2008 the various free market rankings (Heritage/Fraser, etc.) agreed with me, ranking Denmark number one in the world in economic freedom, if you remove the tax/spending category and look at the other 80% of the index (privatization, trade, rule of law, deregulation, etc.)
Heritage now has Denmark #7 in Economic Freedom, far ahead of the US even if you include size of government, where Denmark scores low due to Heritage's small government bias. In areas like business freedom (#1) and property rights (#2) investment freedom and financial freedom Denmark ranks at or near the top of the world:
https://www.heritage.org/index/pages/all-country-scores
Denmark is #9 in the Fraser ranking, but if you take out size of government (taxes and spending) Denmark would be #4 in the Fraser ranking.
https://efotw.org/economic-freedom/dataset?geozone=world&year=2023&page=dataset&min-year=2&max-year=0&filter=0
Obviously these are partly subjective, but note that these high rankings for Denmark are coming not from Bernie Sanders, they are coming from right wing think tanks.
Of course ChatGPT knows nothing about all of this, it's silly to rely on its views.
You’re right. I for one would like to see ChatGPT destroyed entirely forever
I asked Chatgpt the limitations of your viewpoint (just because it might be interesting to you). Answer is here:
3) Important limitations and criticisms
1. It’s not the whole story.
• Banking system collapse, runs, and the resulting destruction of the money multiplier were huge. Those banking dynamics are not just a passive consequence of deflation — they were active drivers of money supply decline. Sumner’s framework can incorporate this (as an amplifier), but it doesn’t by itself explain why banks failed so badly.
2. Direction of causation and multiple interacting forces.
• Was the primary shock a rise in money/gold demand, or was it a collapse in confidence / investment demand that reduced NGDP and raised money demand secondarily? Different interpretations imply different policy priorities.
3. Heterogeneity across countries.
• If gold-demand shifts were the single cause, we’d expect similar patterns everywhere on the gold standard. In fact the depth/duration varied a lot and that variation maps strongly to domestic policy responses (bank rescue policy, fiscal policy, when/how quickly countries left gold). That suggests institutional and policy differences mattered a great deal.
4. Role of real shocks.
• There may have been important real shocks (e.g., productivity reversals in some sectors, sectoral overinvestment from the 1920s) and structural problems (labor market mismatch) that monetary policy alone couldn’t fix. Monetary accommodation helps, but it doesn’t produce investment that’s uneconomic forever.
5. Sticky wages/prices assumption.
• The implied real effects rely on downward nominal rigidity. How sticky were wages and prices across sectors and countries? The degree of stickiness changes the quantitative implications.
6. Debt-deflation and balance-sheet channels.
• Irving Fisher’s debt-deflation theory emphasizes that deflation raises the real burden of nominal debt and bankruptcies, further depressing demand. Sumner’s story is compatible with Fisher’s amplification channel, but debt dynamics are an additional mechanism to model explicitly.
7. Empirical attribution is hard.
- Disentangling how much of the downturn was due to gold-demand driven deflation vs. banking runs vs. collapsing aggregate demand vs. fiscal tightening vs. external shocks requires careful empirical work. Many of these things happened at once.
1st point: "Sumner’s framework can incorporate this (as an amplifier), but it doesn’t by itself explain why banks failed so badly."
Yes, it clearly does. I explain how tight money caused falling NGDP which led to bank failures, as it always does.
The second criticism is apples and oranges. I am explaining why NGDP fell using a gold supply and demand framework. That will obviously causes bearish expectations, falling investment, etc. That's not an "alternative view", it's an implication.
The third point is simply incorrect. I do point out that the varying recoveries depend on differences in domestic policy---especially the timing of leaving the gold standard.
The 4th point is the Austrian argument, which is rightly rejected by most economists. Overinvestment doesn't lead to mass unemployment, as we saw in 2007. It leads to re-allocation from one sector to another. That did not happen in the 1930s, as all sectors fell. The problem was falling NGDP. The same thing happened in late 2008.
5th point is hopeless vague. I pointed out that Hoover and FDR made them stickier than in the early 1920s.
6th Again, I clearly discussed how the deflation led to debt defaults, which caused bank failures and higher gold demand.
7th. Again, All of those points are factors that might influence gold demand. I discussed the factors that I thought were the most important.
Many people overestimate the ability of AI to actually THINK. It's good at rehashing other people's arguments--like a hard working B student--but it cannot actually engage in deep thinking about macro questions. It's not there yet.
4) How convincing is Sumner’s view in practice?
• Qualitatively convincing: It captures an essential truth: the gold standard constrained policy and monetary forces mattered a lot. Monetary contraction (and policy failure to stabilize nominal spending) is central to most modern accounts.
• Quantitatively uncertain: Whether the initial trigger was mainly a shift in gold demand (in the narrow sense) versus banking system collapse, or a collapse in confidence that raised money demand, is debated. Sumner gives a coherent mechanism that complements the Friedman-Schwartz monetary history and Fisher’s debt-deflation story, but it should be read as one piece of a multi-factor explanation.
• Policy takeaway (robust): Don’t let nominal spending collapse. That’s a robust implication — whether the trigger is gold demand, banking runs, or collapsing confidence, stabilizing nominal GDP or satisfying money demand avoids the worst outcomes.
(end)
I think some, or a lot of this, is classic Chatgpt trying to be very neutral on a debated issue. Still interesting though!
OT but interesting. The Swiss National Bank, that is Switzerland's central bank, has a balance sheet equal to 100% of GDP. They are on the cusp of deflation.
Same thing happened to the Bank of Japan, pre-COVID.
Huge balance sheet expansion, and then deflation.
Sigh . . .
Yi Wen, 2013. "Evaluating unconventional monetary policies -why aren’t they more effective?," Working Papers 2013-028, Federal Reserve Bank of St. Louis.
Handle: RePEc:fip:fedlwp:2013-028
DOI: 10.20955/wp.2013.028
OK, if I am nuts, then Yi Wen is nuts too.
Yi Wen doesn't leave dumb comments in this blog.
Wen said QE (LSAPs) would have had to have been four times larger to have had much effect on inflation.
So, I pose the "dumb" question: If LSAPs do not budge inflation much...why not? Why not build the Fed balance sheet, and alleviate taxpayer debts?
My take is with our two-party "system," the US will see federal budget deficits for as far as the eye can see.
The only practical option ahead is to run some inflation to deflate the debt, and then some QE to switch debt ownership back to taxpayers.
Or, we can moralize about the virtue of balanced budgets.
To take this more serious than is probably warranted:
(1) If your money printing isn't causing (enough) inflation, you need to print more money. You can buy arbitrary assets with the newly printed money. Repeat until you reach the desired level of inflation. (Use market forecasts to guide you. Don't use a rearview mirror of measured inflation.)
(2) That has nothing to do with inflating away government debt. The goal in (1) is to get back up to the targeted path of eg 2% inflation, or for the Swiss to hit their targeted exchange rate with the Euro. Nothing beyond that.
I disagree to some extent.
Yes, QE, or LSAP, but also money-financed fiscal programs can be used to fight deflation or sluggish growth.
Plain ol' deficit spending, apparently backed into the US cake for the foreseeable future, can somewhat work (with monetary accommodation) but ramps up taxpayer indebtedness.
The US can avoid over-leveraging by buying back debt, through the Fed.
David Beckworth advocates QE plus tax cuts on ordinary people (QE for Main Street).
My idea is Social Security tax holidays while the Fed buys Treasuries and puts them into the SS fund. This might be a good idea anyway.
Inflation in the 2% to 3% range is OK, even maybe 3% as it helps deleverage the nation.
I don't see the moral or practical superiority of only expanding the money supply through regulated commercial bank lending.
Also, in a world of globalized capital markets...what does a lone
central bank tightening or loosening actually do?
(Maybe in a perfect world, we can balance the US budget. It happened back with Clinton. But realistically, we must devise monetary policy given chronic federal deficits and globalized capital markets, and what is actually in US interests.
If global investors accept US cash for Treasuries, and that has little or no effect on inflation, then fine, let them hold US dollars. Keep US dollars in offshore bank accounts or briefcases. Who cares?)
The GD was caused by the massive decline in of the money supply due to the collapse of the banking system. In other words, by "market discipline".
I'd put it this way:
Tight money caused the Great Depression in late 1929.
The Great Depression caused the banking crisis in 1931.
The banking crisis worsened the Great Depression.
If the Bank of the United States had not defaulted upon its deposits, the national banking run would not have happened:
“There is nothing in the economic situation as it stood in, say, September or October 1930 that made the continued and drastic decline of the following years inevitable or even highly probable. The character of the contraction changed drastically in November 1930, when a series of bank failures led to widespread runs on banks, which its to say attempts by depositors to convert deposits into currency. The contagion spread from one part of the country to another and reached a climax with the failure on December 11, 1930 of the Bank of the United States. This failure was critical not only because the bank was one of the largest in the country, with over $200 million in deposits, but also because , though an ordinary commercial bank, it’s name had led many at home and even more abroad to regard it as somehow an official bank. Prior to October 1930 there had been no sign of a liquidity crisis or any loss of confidence in banks. From this time on, the economy was plagued by recurrent liquidity crises (which) were important because of their effect on the money supply. The Reserve System failed this test miserably. It did little or nothing to provide the banking system with liquidity, apparently regarding the bank closings as calling for no special action. The system had ample power to provide the banks with the cash their depositors were demanding. Had this been done, the bank closings would have been cut short and the monetary debacle avoided. ” (Friedman, Capitalism and Freedom, 1962)
"The Federal Reserve had the power at least to ameliorate the problems of the banks. For example, the Fed could have been more aggressive in lending cash to banks (taking their loans and other investments as collateral), or it could have simply put more cash in circulation. Either action would have made it easier for banks to obtain the cash necessary to pay off depositors, which might have stopped bank runs before they resulted in bank closings and failures. Indeed, a central element of the Federal Reserve's original mission had been to provide just this type of assistance to the banking system. The Fed's failure to fulfill its mission was, again, largely the result of the economic theories held by the Federal Reserve leadership. Many Fed officials appeared to subscribe to the infamous "liquidationist" thesis of Treasury Secretary Andrew Mellon, who argued that weeding out "weak" banks was a harsh but necessary prerequisite to the recovery of the banking system. Moreover, most of the failing banks were relatively small and not members of the Federal Reserve System, making their fate of less interest to the policymakers. In the end, Fed officials decided not to intervene in the banking crisis, contributing once again to the precipitous fall in the money supply." (Bernanke, H. Parker Willis Lecture in Economic Policy, Washington and Lee University, Lexington, Virginia, March 2, 2004, "Money, Gold, and the Great Depression".)
I'd encourage you to read The Midas Paradox, which I believe improves on the excellent work of Friedman & Schwartz and also Bernanke.
OK. I'Ve been meaning to but that kind of book can be daunting.
The Midas Paradox is pretty approachable, to be honest. Not much math in it even.
Is it crazy that I like to regard inflation as a tax on money hoarding? (Maybe "penalty" is a better word than "tax")
This only works if monetary policy is sensible--i.e. not Weimar--and leaning towards stable NGDP growth.
Yes, it is a tax. While the reduction of money hoarding is a valid objective, the cost of high inflation still exceeds the benefit. But it does make the optimal rate of inflation a bit higher than what you'd get in simple models that ignored tax evasion in the underground economy.
Well, anticipated inflation isn't much of a tax, is it? Or rather, it's easy to evade by holding less cash?
The biggest effect in a modern economy is probably from its interaction with capital gains taxes?
I'm thrilled to hear you are attracting many new readers. I hope you will take that opportunity to tout (and they will read) your online book Alternative Approaches to Monetary Policy!
At some point I'll post on that book.
Would be interested in your view of Gauti's latest https://www.aeaweb.org/articles?id=10.1257/jel.20241306
Haven't read the paper, but here are a few comments on the abstract:
Conventional monetary policy is ineffective at the zero bound only if you define policy as interest rate targeting. But I don't find that to be a useful definition of monetary policy.
FDR did not do much fiscal stimulus; the recovery was largely caused by monetary stimulus.
Price fixing policies such as the NIRA were contractionary, and slowed the recovery.
This seems to suggest that if nothing is done, wages and prices will eventually correct to the new equilibrium, and no intervention is required at all. But in this context, that means deflation, and deflation incentivizes hoarding currency, which produces a feedback loop of reduced purchasing -> reduced demand for goods -> prices fall further. This has to bottom out eventually, of course, but in the mean time a lot of benefits from exchange are lost.
"This seems to suggest that if nothing is done, wages and prices will eventually correct to the new equilibrium, and no intervention is required at all."
It depends what you mean by required. Even if wages and prices are flexible in the long run, you can have a lot of suffering in the short run. So you'd still prefer a monetary policy that produced stable NGDP growth.
This response along with your other response about inflation being a tax, leads me to a question: If I remember correctly from past posts on econlog, you favor a NGDP growth target of about 4%? If RGDP growth is 2% long run, why not a 2% NGDP target, netting to 0% inflation in the long run? Is it just to target tax evasion in the underground market? Does that really matter? And would it really need to be another 2% (assuming I remember correctly that you would target 4%) or would something like a 2.5% NGDP growth target be enough, while also keeping long term inflation low?
Mainly because nominal wages are inflexible in a downward direction due to money illusion. Workers don't like negative nominal wage changes. In addition, with 4% NGDP growth (level targeting) you don't have to worry about the zero lower bound for interest rates.
I am not sure I understand the situation the first would matter (the latter makes sense). What is the circumstance in which there would be the need for negative nominal wage changes if NGDP was growing at 2%?
If RGDPg is negative (or lower than 2%), there will be some inflation to get NGDPg to 2%, so there would be no need for negative wages. If the economy is running hot, with RGDPg greater than 2% then the deflation would be in sectors other than labor. At the worst, holding labor costs constant would cool the market off, bringing it back down to the 2%.
I am probably missing something here, I realize that, it just isn't clear to me what it is.
Think of NGDP as driving the average wage change. Actual wage changes in various sectors look something like a bell-shaped curve, as each sector faces different supply and demand shocks. Unfortunately, the bell shaped curve for wage changes is truncated at zero percent, so sectors that need to see their relative wage fall have trouble making adjustments when average wage growth is quite slow.
You would be correct if all sectors had the same equilibrium nominal wage increase.
That makes sense, obviously it would still happen at 4%, but the tail would be much smaller. What is the standard deviation of this bell curve, is the tail past 2% really large enough to be problematic (if SD is less like .5%, I wouldn't be concerned with the >4SD industries; if it more like 1.5%, I see the problem)?
However, it does seem like some industries have figured this out on their own. At the last few places I have worked, 10% of my base salary is an expected bonus. The actual number can vary in both directions depending on company metrics. Like I am expecting a bonus of about 9% for 2025 (I will receive it in March). In past years it has been as high as 12%. This gives them the ability to give standard raises every year and still have negative wage growth in down years (If I get a 3% raise but only get a 5% bonus, I am making less). I realize this isn't possible in every industry, but its a free market solution to that issue.
If you like these kinds of speculation, you should check out George Selgin's 'Less Than Zero' (https://cdn.mises.org/Less%20than%20Zero%20The%20Case%20for%20a%20Falling%20Price%20Level%20in%20a%20Growing%20Economy_4.pdf)
In more 'Sumnerian' terms George Selgin explores an economy with a 0% NGDP growth target (or rather a flat NGDP target level).
Our dear host Scott Sumner also wrote the foreword to the 2018 edition of Selgin's book.
I’m interested in that claim that the central banks held most of the gold that had ever been mined. I’ve heard in various contexts that households and temples in India actually contain a vast majority of the gold that has been ever been mined. Do you know if there’s any actually good measurement of this? (It’s definitely hard to know what is actually in the vaults below the temples in India.)
I don't recall the source, but we had good data on total central bank holdings, so it's a question of how much gold has been mined over time. Mining output rose a lot over time, so I suspect that a good portion of the mined gold was produced between 1800 and 1930, for which we'd have reasonable data. But that's just a guess.