The economics of supply cut-offs

As a number of people have pointed out, cable TV, cable internet connections, and cable streaming have been remarkably robust throughout this crisis.  Why might that be?  Let’s think through a few basic points about the economics of supply cut-offs.  This will not be a complete model, but it will focus attention on perhaps one possible factor of interest.

Imagine a seller with market power who comes close to perfect price discrimination.  That supplier will take great care to avoid supply cut-offs (imagine an electric utility investing in emergency capacity, for instance).  If a cut-off were to happen, the profits of the supplier would be much lower.  As a first-order approximation, such suppliers will invest a near-optimal amount of resources to prevent such supply interruptions.

Alternatively, imagine a nearly perfectly competitive situation where all of the surplus goes to consumers and producers earn the going rate of return.  Fixed costs are not significant.  A market collapse or supply cut-off doesn’t cut much into profits, and in essence the suppliers do not care about the losses of the inframarginal consumers, were a supply interruption to occur.

As a simple theorem, if the market is good for the producers in the first place, supply interruptions are less likely.  If the market is good for consumers in the first place, supply interruptions are more likely.

Might this also apply to health care systems?  The U.S. hospital system, in normal times, spends way too much.  Still, it has the “cultural mentality” for making capital expenditures, far more than say Britain’s NHS does.  And so the United States has far more ICU units per capita than does Britain.  Whether justly or not, the U.S. health care system might come out of this crisis looking not entirely bad.


"Still, it [the US hospital system] has the “cultural mentality” for making capital expenditures, far more than say Britain’s NHS does."

In other words, the US [inter-temporal] budget constraint is looser. :-)

'Why might that be? '

This is the sort of highpoint that makes MR stand out from the crowd - someone as knowledgeable as TC explaining technology and why something like a cable connection still works after a couple of weeks of a pandemic spreading in some regions of the U.S. Not electricity, which is boring and gets cut off on a regular basis in much of the U.S., but the really important things in life.

However, the NHS is most certainly not cutting the pay of its medical workers in the middle of a pandemic, so the U.S. health care system might come out of this crisis still looking entirely exceptional.

Cancelling elective procedures was done to add capacity to deal with the pandemic. It doesn't seem to have been needed, at least yet. In the meantime there isn't enough for all the people and resources available to do, and that is a testament to the amount of capacity there is available. That is the point the Tyler is making.

Whether it is being used appropriately is another question.

Of cutting the fat, American health care style. This quote really needs to be emphasized in terms that at least some resources are being cut that do involve treating pandemic patients, "Hospital executives and analysts emphasize that not all the furloughed or fired workers are directly involved in treating covid-19 patients."

From yesterday, technically -
Hospitals across the country have deferred or canceled non-urgent surgeries to free up bed space and equipment for covid-19 patients. But that triage maneuver cut off a main source of income, causing huge losses that have forced some hospitals to let go of health-care workers as they struggle to treat infected patients.

Last week, Bon Secours Mercy Health, which runs 51 hospitals in seven states, announced it would furlough 700 workers. On Wednesday, Ballad Health, which operates 21 hospitals across Tennessee and southwest Virginia, delivered the same bad news to 1,300 employees and said executives would take pay cuts. Employees at Children’s National Hospital in the District were informed this week that they must take off one week, using either vacation time or, if they have none, unpaid leave.

Hospital executives and analysts emphasize that not all the furloughed or fired workers are directly involved in treating covid-19 patients. Others say the furloughs help reduce the number of people in hospitals, slowing the spread of the virus.

How many analysts are following the NHS market segment, and when did it become considered completely normal to think that analysts involved in share prices or profit forecasts are an apparently integral part of the American health care ecosystem? Are there undiscovered synergies requiring thinking outside of the box to deliver greater share holder value in the form of executive bonuses?

Where the hell in the U.S. do you live that electricity "regularly" gets cut off?

Anywhere that normally has tornadoes, thunderstorms, blizzards/ice storms, or hurricanes. It is completely normal to read about how power has gone out for tens or hundreds of thousands (and more than occasionally millions) of customers on the East or Gulf Coasts, along with other regions that also have extreme weather.

The US is big, so a state like Washington might not fit into that description.

Dave, questions like that are just feeding the troll. Notice that he immediately and tacitly abandoned his initial suggestion that electricity gets cut off more than cable, and retreated instead into irrelevant pedantry. It's the same way with his flip-flops along the way about US hospital capacity (he used to predict that it would be overwhelmed). This isn't someone arguing in good faith.

"Where the hell in the U.S. do you live that electricity "regularly" gets cut off?"

prior, says he lives in Germany and has for decades. It's just his stick to poke at America at every opportunity.

I'm sure he's basing his comment off the statistic that:

The grid reliability between the US and Germany
US at 99.96% vs Germany at 99.997%

And Germans only pay a bit under 3 times what Americans do per kWh for that extra reliability.

My cellular internet has been way better than my crawling cable connection. And if there were a competitive market which allowed for price gauging in emergencies, companies would be just as interested in having emergency backup.

doesnt explain germany's surplus afaik. presumavly there are other ways to get there..

Germany, with its for-profit, yet more or less single payer health care system is not a model for the U.S. Don't ask why, just watch the American version of Upper Class Twit of the Year instead.

Regardless of how good US health care system is- do we really know how effective treatment in a hospital is anyway? I mean does it make you 80% less likely to die or 40% or 5%? Anyone know?

What I undesrtood from different sources is that it is very different depending on the patient and the kind of treatment. A patient with low blood oxygen will probably be saved by a standard oxygen mask. At the stage when ICU is needed, however, the survival rate drops also with proper treatment (in the UK they say around 50%). However, a lot depends on the general condition of the patient: a 80 years old with chronic heart problems put in ICU needs a miracle, an otherwise healthy 60 years old is likely to make it through.

Thanks. I guess if you need ICU you would almost certainly not survive without it. So if 50% make it after that step- well that is effective treatment in my estimation.

Something of a matter of definition. There are ICU beds without ventilators or where they are not being used, such as with Boris Johnson. Apart from supporting breathing, there is no treatment by a normal definition either - you either recover or you don't. Breathing support is provided to allow more time for a recovery to occur.

In the US CDC report 2449 cases:
26% hospitalized
8.2% needed ICU
2.6% died

On a slightly different metric, the United States ranks 32 in hospital beds per per capita which is near the bottom for OECD (but it does beat the UK at 35). Japan, South Korea, Russia, Germany, and Austria, in that order, top the list. Depending on how you define "supply cut-off", you could argue that the US is one of the best (ICUs) or one of the worst (hospital beds).

So much of what they do is out patient anymore. Better results.

Certainly, plenty of people were predicting recently that US hospital bed capacity would be overwhelmed. It will be interesting to see if anyone ever examines the reasons for the inaccuracy of these predictions, or if they will just be memory-holed.

Memory-holed. the inaccuracy is inconvenient to the idea that we need single-payer in the USA.

Yes, but another metric is “ critical care beds per capita” . I assume that means ICU beds.
Here it says the US ranks first.

The US has more beds, but has a normally higher occupancy rate for such beds.

The two tables do not represent the same beds. The second one includes less intensive care beds ( IMCU). The US occupancy rate for ICU in for normal times is ~ 2/3, The US has more ICU beds available, Japan has more ( ICU+ IMCU) beds available. These IMCU might just be average beds without negative pressure etc..
Otherwise the math doesn't work in this article ( if they are the same beds)
Table 1 says the US has 25.85/7.9 = 3,27 more ICU beds than Japan ( per capita)
Table 2 says Japan has 19.1/8,8 = 2.17 more beds available than the US due to occupancy rates
This implies the unoccupied bed rate in Japan is 7.1 that of the US.
Most studies give the US unoccupied bed rate at 1/3. This would give the Japan unoccupied bed rate at 2.33 ( Greater than 1!).
Regardless, I would be interested in knowing ( pre-Covid-19), how the US compares with other countries as to the ICU admissions per capita.

Maybe they are robust because internet connections are incredibly scalable? Most of the internet pipe is used for video streaming (in one way or another) and the difference between SD and HD (1080, 60fps) bitrates are almost 3 times. And you can go lower. That's exactly why youtube has made SD option the default in recent weeks.

I understand how the argument works for electricity or mobile internet, where you pay as you use, but how does it apply to broadband - where you are charged a fixed sum per month? Surely, a seller with great market power would not care much about interruptions, as the income remains unaffected and your clients have no alternative suppliers to switch to? 🙂

+1. Hospitals are paid per use. Broadband is an all you can eat buffet at a flat rate. Even though both are local monopolies with the accompanied regulations, I'm not confident the analogy works.

I think most people pay a flat rate for mobile internet as well. And mobile is increasingly a competitor of fixed broadband.

A different way of trying to explain the phenomenon is that increased market power permits a supplier to amass sufficient resources to prepare for long-term risks. In competitive markets, where rents are competed away and only razor-thin profits accrued to efficient players are possible, any peacetime investment in outside risk mitigation appears to be an unaffordable extravagance.

This feels like trying too hard...

Could it just be that bandwidth wasn’t a binding constraint to begin with? Certainly for the backbone infrastructure there is a ton of excess capacity built in. But is this because of monopoly power, or just a high fixed cost of adding capacity, desire for scalability, and some regulatory pressure? Build a lot now so the marginal cost of adding new customers is low, but this also helps when existing user base starts using network more.

For the monopoly story to really bite you’d have to think that customers can refuse to pay if service deteriorated, or that people are paying per “unit” delivered. In that case we need to keep delivering “units” to take in the high rents. That’s not how the system is set up. Lots of store renters would love to stop paying right now since they’re not using those “lease services” but that’s not how the contract is written. Same with your ISP - you pay no matter what during the life (1-2 years) of the contract. If my internet service suddenly got worse I’m still on the hook to pay, and right now the no-internet outside option isn’t look so good. If anything my willingness to pay for even the proverbial “few drops” of internet is very high at the moment so I’m not in much of a position to argue over connection speeds. Speed degrading shenanigans have been documented in the past (fights with Netflix) but no bills were affected or refunded to my knowledge.

Another technological explanation is that the flexibility to transfer resources between commercial and residential are easier technologically for this sector than it seems to be for toilet paper. To what extent is the infrastructure even segregated beyond the “line to the house”? If anything I would guess that total usage (residential + commercial) is down, and the service standard is certainly lower for residential (nobody expects home internet to be as good as the higher end corporate plans). Easier to deliver on a worse promise.

Ooh, ooh, ooh; we want to learn how to defend ourselves against socialized medicine, do we? Getting all high and mighty, eh? Fresh pandemics not good enough for you, eh? Well, let me tell you something, my lad! When you're walking home tonight and some great homicidal maniac comes after YOU with a bunch of nvel coronavirus, don't come cryin' to me!

Which seems fitting connection with something basically impossible to imagine happening in a single payer system.

“I’m concerned about the [financially] weakest 25 percent of hospitals, because there’s no way the other hospitals can absorb their covid patients,” said John Fox, the CEO of Beaumont Health, Michigan’s largest hospital system.

“If they start going down, that changes the whole algebra for the size of the system to handle the pandemic,” Fox added. He worries that a surge in patients could overwhelm more stable hospitals. “That’s when you’re treating people on the front lawn.”

It has also made it harder for some hospitals to finance the considerable costs of treating covid-19 patients — from efforts to purchase personal protective equipment and ventilators to setting up drive-by testing sites.

A report this week by the Inspector General for the Health and Human Services Department blamed the cash crunch for ill-timed layoffs and furloughs.

“Hospitals reported laying off staff due to financial difficulties, which further exacerbated workforce shortages and the hospitals’ ability to care for covid-19 patients and the routine patient population,” the IG found.

“One administrator stated that it had been ‘an absolute financial nightmare for hospitals.’ ”

In contrast, the NHS is being provided as much financial support as possible by the UK, upper class twits running things or not.

They may not be treating people on the front lawn in Michigan yet, but only because people are dying in hospital hallways instead. "We've had patients die in hallway beds because the nurse didn't find they didn't have a pulse until it was too late," the physician told CNN, speaking on the condition of anonymity. "Each nurse has so many patients that by the time they come to check on their next one, there is a chance that patient may not have a pulse anymore."

Detroit has the most cases in the state, with 5,824 and 251 deaths, and the state has the third-highest number of cases in the country.

Yes, why can't the U.S. have a health care system like the NHS? All of that state support and government planning means that the NHS has plenty of ventilators to equip their ICUs to meet peak demand.

Or not:

Yet the UK (for a number of reasons) does not have lists like this.

For the analysis, Kaiser Health News identified the number of ICU beds each hospital reported in its most recent financial cost report, filed annually to CMS, and then totaled the number of ICU beds per county. The ICU bed tally does not include Veterans Affairs hospitals.

U.S. states ranked by the number of counties with zero ICU beds:
Note: The list includes ties and results in a numerical listing of 39.
1. Texas: 166 counties
2. Georgia: 95 counties
3. Nebraska: 81 counties
4. Missouri: 78 counties
5. Kansas: 76 counties
Virginia: 76 counties

Virginia and Georgia are not exactly population poor western states with large amounts of empty land. And at least part of Virginia is expected to be the next hot spot, though one can imagine that Northern Virginia, which one could believe is as well equipped with hospitals/ICUS/ventilators as the NYC region, will do as well as NY in dealing with a pandemic.

Northern Italy, the richest region of that country, is about as well equipped as the U.S. on average. As the county list demonstrates, averages can be a bit deceiving. And the link below shows how completely deceptive averages are in Virginia, where a notably large amount of the state has counties with no hospitals at all.

Particularly when looking at this link - More than half of counties have no hospital ICU beds, a growing concern as the novel coronavirus spreads throughout the nation. This map shows counties with no hospitals, counties with hospitals but no ICU beds, and counties that do have ICU beds. You can scroll over the map or, better yet, do a more comprehensive search on KHN’s interactive table, where you can search by either county or state.

Part of your list appears just to be "U.S. states that have a lot of counties".

Texas has 254 counties, for example (the most of any U.S. state). Many of them are very low population because, even in the sparsely populated parts of West Texas, the counties are not that big in land area. My understanding is that goes back to an idea when the county lines were drawn in the 19th century about making it relatively easy to travel to the county seat in 1-day on horseback.

That also means that, with modern methods of transportation, having to go one or two counties over to reach a hospital isn't really that bad by the standards of what one should reasonably expect for a rural area or very small town.

That ambulance ride from rural areas must be the mother of all hospital bills though.

To repeat, neither Georgia (2) nor Virginia (tied for 5) are low population western states, and their county lines (certainly Virginia's) basically predate the existence of the U.S. entirely.

As for that ambulance ride - undoubtedly every ambulance is equipped with a ventilator, and the personnel to man it, since that will be an important criteria during the current pandemic if someone is in an ambulance for a longer period of time.

In Virginia, any self governing entity (e.g. the City of Falls Church) is listed as its own "county". Things makes Virginia the state with the third most counties in the country at 133. Coupling that with the large swathe of thinly populated counties down in the Southwest likely explains pretty much all the variance.

Georgia has 159 counties. The second most in the country. They also have a large number of sparsely populated counties up in Appalachia.

The only surprising thing about your list is that Kentucky, Illinois, and North Carolina are not on it. Otherwise it is just which states have the most counties with a bit of random noise.

The troll is only pretending not to understand the difference between urban and rural areas within a state.

Yet using the map from the second link, it is the broad middle that is without hospitals at all, not the southwest of Virginia.

That list includes
7 Kentucky: 67 counties
8 Illinois: 56 counties

Surprisingly, North Carolina is tied at 19 with Michigan

The link is

The broad middle of VA tends to be within an hour or so of the major hospital centers - Blacksburg, Charlottesville, Richmond, Lynchburg, etc.

It is also relatively low population.

Seriously, who creates a health care system that is apparently designed to financially and then functionally collapse in the face of a fresh pandemic attack? "These experiences reflect the losses across the entire health-care system given the cancellation of nonessential procedures, analysts said. A typical hospital system with 1,000 beds and the ability to perform outpatient surgeries is predicted to lose around $140 million — half its operating revenue — over a three-month period, the Advisory Board, a consulting firm, reported this week.

Further complicating problems has been the need to finance the purchase of additional equipment to protect staff and prepare for coronavirus patients."

What happened to the flexibility of the market to react to changing circumstances, in contrast to the clay feet of anything associated with government? 'Sorry squire, I've had a look 'round the back of the shop, and uh, we're right out of hospitals.'

Based on estimated CFR, deaths per million population, or peak increases per million in death rate, the more market oriented systems (notably the US and Germany) have been doing vastly better than the more centralized ones (Italy, France, Spain, the UK, and the Netherlands).

Maybe it is all luck, but for the metrics I care about the most, the US system has been performing better than average in the Western World. Maybe this will all turn out to be artifacts once we do the sort of deconfounding that is necessary for intelligent international comparisons, but for the first order comparison on the most relevant measures the US is doing well. Excluding NYC, which is of course the market which most closely resembles a socialized government system in the US, and we are doing remarkedly well.

NY's (including NYC) death total per capita today is higher than Italy's, while the absolute number of cases is higher.

And do not make too many predictions about the future yet - the Italian and Spanish lockdowns are considerably stricter than those in the U.S., but the (encouraging) effects take weeks to become a trend that shows improvement - from a high number, however.

" The U.S. hospital system, in normal times, spends way too much. "

In abnormal times, they spend too much too. That's why doctors and nurses are now taking pay cuts or getting laid off. Let's not pretend a $130 billion bailout specifically for hospitals didn't just happen either.

But look at the details of the boomer hospital bailout. Though undoubtedly how it works is yet another one of those coincidences that happen so often in these days of bailouts. "Verma said hospitals would receive the first tranche of aid based on the volume of services they bill for Medicare patients. “This is not on a first-come, first-serve basis,” she said.

But that could leave hospitals that treat more people on Medicaid, or who have no insurance, without an immediate boost when they are hemorrhaging cash.

The approach “could tilt the playing field against” hospitals that primarily serve low-income and uninsured patients, Bruce Siegel, the president and CEO of America’s Essential Hospitals, said in a statement. “It is especially concerning because their narrow margins and front-line role mean these hospitals are among those in greatest need of funding support.”

Amusing that the bailout process almost seems designed to cause the further financial collapse of further parts of the American medical system right in the middle of a pandemic. There is no way that the U.S. health care system comes out of this crisis not looking entirely bad, if there is any justice.

The virus overwhelms even those who try to prepare for it. "New York makes mass graves

New York City officials have hired contract laborers to bury the rising number of dead in its potter’s field on Hart Island, an area which has for decades been used to bury those with no known next of kin, writes my colleague Joan E Greve.

New York broke its record for the largest single-day coronavirus death toll for the third consecutive day, Governor Andrew Cuomo announced on Thursday, as he warned the effect of the outbreak on the state’s economy is expected to be more devastating than 9/11.

New York recorded 799 deaths from coronavirus on Wednesday, bringing the state’s total death toll to 7,067. New York has lost about the same number of people to coronavirus as the UK."

Not sure this situation counts as a supply shut off in a precise sense, but most major cities such as Madrid or London or Paris seem to be able to create temporary morgues without resorting to burying the dead in a park.

From the Washington Post, a few more details.

As the number of people dying in New York each day reaches grim levels, city officials have hired contract workers to bury the dead on Hart Island, a longtime burial ground for the city’s indigent residents.

The island, a one-mile strip off the Bronx accessible only by boat, serves as the final resting place for more than a million former New Yorkers — those who died without known next of kin or whose families could not afford a funeral.

Normally, about 25 corpses are buried on the island each week by low-paid inmates at Riker’s Island, the city’s notorious jail complex. But with a fivefold increase in bodies and a virus outbreak at the jail, the city has turned to contract employees instead.

Jason Kersten, a spokesman for the city’s Department of Correction, told Reuters that about two dozen people are now being buried daily, five times a week. Two new trenches have been excavated by digging machines to accommodate the uptick in deaths.

But do note the detail that 'uptick' is used to describe a fivefold daily increase in burials.

For those without reading comprehension issues, the detail described an uptick in deaths, not burials.

But, As the number of people dying in New York each day reaches grim levels, many people die of all causes in New York on an average day? These corpses being buried on Hart Island, they have no next of kin? They're indigent? Or the croup has overwhelmed not only medical facilities but also mortuary and burial services? By the way, most graves are dug now by digging machines, not convicts with shovels. And, what of the crematorium business? Is it booming or simply so overwhelmed with demand that they've given up?

'For those without reading comprehension issues'

Um, it is a five fold increase in burials, from five a day to the current two dozen a day. "Normally, about 25 corpses are buried on the island each week by low-paid inmates at Riker’s Island, the city’s notorious jail complex. But with a fivefold increase in bodies and a virus outbreak at the jail, the city has turned to contract employees instead.

Jason Kersten, a spokesman for the city’s Department of Correction, told Reuters that about two dozen people are now being buried daily, five times a week. Two new trenches have been excavated by digging machines to accommodate the uptick in deaths."

It would appear that to increase productivity, with a commensurate decline in wage costs, the graves are being dug using equipment beyond shovels in the hands of prisoners.

The cable providers, and data providers have networks that can handle peak hours. ( use to be 7 to 11pm ). I don’t think Covid-19 has resulted in much higher use than that.
If the system is overloaded, it’s a soft failure. Most people connections slow down but still work, video resolution is downgraded some and it is not perceived as an outage by the user. I have noticed more slowness in internet use in my area.
Are the high costs of health care in the US due to capital expenditures ? These should to some extend deflate due to Moore’s law.
I think the high costs come primarily from high salaries of doctors and health care workers , a high ratio of specialists to generalists, and the administrative and regulatory burden.
The US has one of the highest number of critical care beds per capita in the world and I read that the occupancy rate is about 2/3. It has 5 times more then the UK or Japan.
What does this mean ? Americans are hospitalized at much higher rate or there’s a chronic shortage in other countries ?

We have a mathematical proof now which tells us that sometimes the value chain will not converge. But in the case of continuous adaption the value chain will be stable. I think.

I'm missing the perfect, or near-perfect, price discrimination among internet service providers. I suggest the robustness in networks is not remarkable but more a matter of time of peak load moving rather than the amount of peak load. The network does not care when the peak is, just that it not increase greatly. Further out there, I suggest the FCC repealing network neutrality two years ago gave service providers more confidence to invest in infrastructure. They invest for a few years ahead, but maybe recently more so with the added regulatory confidence.

What we have right now is network neutrality, what was repealed was Network Neutrality (TM) - a scheme favoring Google, YouTube, Netflix, Facebook, etc. over consumers.

As a simple theorem, if the market is good for the producers in the first place, supply interruptions are less likely.
The stable condition is the supply disruptions everywhere along the chain are equally probable, that is the Nash equilibrium. At this optimum, every where supply equal demand within the same uncertainty and intermediate nodes in the chain can calculate fixed inventory costs and set price. This is abstract tree theory in action. Notice it is also congestion management. Notice also that at equilibrium, within some bound error the money velocity equations work.

In the real economy, the credit system does this also, acts like a value chain. So one can see that when the Fed distorts the flow, the velocity equations do not work and losses pile up after the final tally.

The corollary here is that there is no such thing as perfect competition because there is no global view, everything is local and everything accepts an error term, we never get perfect equality in supply and demand, not even in passing. This is the fixed, market uncertainty problem, our equivalent to Planks uncertainty.

The other corollary is that ure federal government is a deliberately skewed value net that cannot adapt. Until we find an adaption protocol for the senate and house then we will cycle atrociously and eventually collapse and restructure. This is why the abstract tree theory recommends that the House and Senate engage in one massive cash transfer to state capitals at each budget period. We have to put a trading pit between the two so senate and house can construct the cost of the value chain skew in federal government. Otherwise we will cycle every eight years in a recession, and the new theorem should prove this to be the case.

I wonder if there are parallels between supply cut-offs and excess demand surges. There's a lot of literature on the latter, e.g. how much electrical generating capacity should a market have, to deal with not just standard demand levels but occasional spikes in demand? Departures from marginal cost pricing might be required. Sometimes the market is treated as a natural monopoly or at least highly regulated. Other times, especially if the good can be stored, the country creates a strategic reserve. Or we just rely on the free market to handle the demand surge.

Those have been well-researched areas for decades.

But if supply cut-offs are different from demand spikes, then yeah we need a new set of models and theories.

I wonder if there are parallels between supply cut-offs and excess demand surges.

Equally probable at equilibrium, and hopefully a small probability. In value chain, your demand is my supply.

Think of everyone having a basket, called inventory. All parties want enough room in their basket so they can purchase supplies when readily available. And they want enough inventory in this basket to skip purchases when supplies are not readily available. Everyone is hedged for the same probability of disruption.

The entire universe works this way, it is the math of self sampling systems.

I tire of doing this, and would write the book if I was not a lazy ass hole. The book is coming, the book is derived from the new proof.

But the whole point of making the tree trunk round is to make all inventory shipments look like centered bell shaped curves, Gaussian. At that point, arrivals and departures of inventory are independent events with minimal cross interference (they do all pile up on the dock at the same time). It was economists who discovered this, Hayek and Krugmam agglomeration theory. Then Fama information processor, then Nash,Coase and Shannon. I keep repeating the chain, everyone knows it by now.

So the theory of stable value chains is Hayek, not Keynes.

What does that say about Keynes? He believed, like Newton and Thomas Aquinas that Godot has placed perfect measuring marks all about our economy. If the marks are off, then we can do a stimulus and replace all the marks with more accurate marks. This is the Hicks cross Krugman bandies about.

It is complete horse manure, and the only purpose of the theory was to cover up the sudden default of Roosevelt. It is a 'This time is different' theory and government paid a lot of research money to promulgate the fraud after the 1972 Nixon defaults.

It's the efficiency of the delivery system that is magnified in this time of crisis. The delivery system for cable and internet, once put in place, doesn't require that much labor to deliver the goods. The delivery system for groceries, on the other hand, requires an abundance of labor and is not very efficient. How many days have you had to wait for Instacart to deliver the goods? Indeed, even the delivery system for Amazon, considered highly efficient, turned out not to be so efficient after all: even Prime customers may have to wait a month for delivery. Thus, the distinction between virtual goods (like cable and internet) and tangible goods (like groceries and home or office supplies). If pandemics and lock-downs are our future, will the delivery systems for tangible goods be greatly improved, or will virtual goods become the favored goods for consumption at the expense of a decline in the consumption of tangible goods?

Pandemics magnify the benefits of robots and automation:

If the producer is able to perfectly price discriminate it is also likely they can control any supply cutoff to protect the supply to the higher valued segment of demand at the expense of the less valued segment of the demand. That is really not the same thing as limiting supply cut-offs at the market level.

It probably is true that such a price discriminating entity might use the additional surplus extracted from the higher value end of demand to have a larger capacity for output if they cannot discriminate in supply delivery. However, I am not convinced producers are unable to selectively offer to its preferred segment of demand.

Does anybody have data that internet traffic is higher now that other peak internet demand periods?

A couple of other thoughts:

The visibility of the monopolist to the public and to regulators might provide a strong disincentive to curtail supply during periods of critical need.

Rayward makes another good point, one which might be thought about even deeper. He says that delivery system efficiency is key. I might tweak that and state that delivery system *capacity* is key. Amazon's delivery system was super efficient at expected throughput levels, with package delivery in a single day! The robustness of cable and internet might be partially attributed to prior investment in delivery capacity.

As as economist in the utility industry (electricity), I see this incentive structure as well: building more capacity allows justification for rate increases on one hand, including the value of risk aversion (outage mitigation), and also represents risk aversion by those managing the system who don't want to be seen an incompetent (responsible for outages) even if such outages would be socially cost effective.

Right, electrical utilities have to plan for demand spikes, which might be similar (except opposite signed) to supply cut-offs.

The two questions though that I have are: we've got decades of research into optimal capacity and optimal pricing including non-marginal cost pricing for utilities. But those are highly regulated markets; do we went to treat other supply markets similarly?

And also, are demand spikes and supply cut-offs symmetric, or are there important differences? I can't think of big asymmetries off the top of my head but I suspect there may be some important ones.

Perhaps it is easier to model peak demand for cable/internet than healthcare services. Peak demand for cable services may be in the evening when most are at home which may not change much during a pandemic, so the system is engineered to handle that. Peak demand for healthcare varies wildly from the norm with a pandemic and as we've seen is not build to handle that.

"As a number of people have pointed out, cable TV, cable internet connections, and cable streaming have been remarkably robust throughout this crisis. Why might that be?"

1. These businesses have high capital costs but near-zero marginal costs. Therefore cutting off customers who don't pay but may do so in the future would be a big mistake, as it costs more to cut off free riders than to let them ride. (Electric utilities also have very high fixed costs, as well as regulatory restraints).

2. Paying for Cable, Internet or streaming services remains a negative-option for practically all customers: the service providers just keep charging your charge card unless/until you tell them not to (or until the charge is declined).

This is the next best thing to a guaranteed revenue stream, but (fortunately) few vendors can get away with doing this: or most, customers must make a conscious decision to continue buying your product or service.

BUT, why are Corvid-19 statistics presented in popular media only as cumulative cases and cumulative deaths? Instead of percent-cases/deaths-per-1000 in the last day/week/month? The latter provides a snapshot showing relative rates by geography, plus a time series that is easily interpreted as "it's getting worse/getting better." Whereas cumulative statistics can never decrease, but at best stop increasing.

It's worth noting that for many internet companies, the current scale isn't that far off previous peaks. These companies are used to spikes

Amazon: Black Friday
Steam: major game releases
Netflix: new TV series

Best article which is worth reading it .Thanks for info

Comments for this post are closed