Can the US Constitution be repaired?

This is an abridged version of an essay on Medium. 

So many people agree that it is impossible to pass any amendment, that the short answer is "No".  Nevertheless, here are some top priorities, in any case.

The Consitution is better than it's predecessor, but that's not saying a lot.  The Articles of Confederation, which defined the US government from 1779 to 1787, required unanimous agreement from all 13 states on any changes.  The Constitution that we've had since then requires "only" a two-thirds vote in both houses of Congress, followed by ratification, which requires approval from three-fourths of the states (38 out of 50).  

The most recent amendment, the twenty-seventh, which prevented Congressional salary changes from taking effect until after then next election of the House of Representatives, was ratified in 1992, 34 years ago.  The Equal Rights Amendment was introduced in 1923 and passed by Congress 49 years later; after 103 years it has still not been ratified.  The amendment process is clearly deadlocked with no obvious way out at this time. But times can change.  So just to get some directions on the record, here are the four most important things that need to be fixed.

Unlock the Supreme Court

The Supreme Court is the apotheosis of the excessively powerful, unaccountable "unelected government official" that conservatives used to decry.  The Constitution gives them "supreme" power to reinterpret laws passed by Congress and veto actions of the Executive Branch, with no further recourse.  Individual justices serve "during good behavior," which has been interpreted since the beginning as life tenure, revocable only by retirement or impeachment. The court, self-servingly, has had no objection.

There have been many proposals for limited terms for Supreme Court justices.Thomas Jefferson, for example, denounced life tenure as wholly inconsistent with our ordered republic. Accordingly, he proposed renewable terms of four or six years for federal judges. Members of the Board of Governors of the Federal Reserve System serve for fourteen years. Eighteen seems sufficient for any Judge or Justice.

The goal of term limits is to make the court more reslisient against ideological and political influence and ossification by refreshing its membership from time to time. That resilience could be made even stronger by electing justices from appellate courts and by those appellate justices rather than allowing the president to nominate anyone he pleases, eliminating the opportunity for an unscrupulous president to nominate inexperienced ideologues.  Stability arising from diversity of experience could be further enhanced by rotating candidates from among each Circuit Court of Appeals.

Rebalance Senate Representation

The United States originated as a federation of independent colonies.  In order to motivate those colonies to join together into a single nation, they were given separate representation in a Senate with equal representaiton for each state.  The distinct character and interests of each state was ensured by the diffuculty of transportation across interstate distances.  Two hundred and fifty years later, technology has made the natural size of a cultural and economic unit far larger: the megalopolis between Boston and Washington operates in a unified way that makes the eight states that it spans into largely irrelevant details.

Representation in the Senate needs to be restructured on the principle of "one person, one vote." The simplest change would be to eliminate the Senate entirely, but this would also remove the stability provided by a second smaller body with longer terms of office than a huge single-chamber Congress of two-year members.  Keeping the Senate, but reorienting it to represent people rather than land could be accomplished by providing one senator for each million people, and retaining the minimum of two senators per state, and retaining the six-year terms.  With a US poopulation of 330 million, that would give a much larger Senate than we have now, with about 300 Senators.  California would have 39, and Texas would have 32 Senators.

I'm attracted to a more radical redesign, that would retain the balance of a large House of Representatives with short terms of office, and a smaller Senate with longer terms.  This design would designate thirteen regions of equal population, in memory of the thirteen original states. They could be based roughly on the "American nations" research of Colin Woodward, with the addition of representation for native americans in the former Indian Territory of Oklahoma, Navajoland in Arizona, and the Pueblos of New Mexico, and representation for non-state territories such as Washington DC, Puerto Rico, and the pacific island territories.The establishment of regions larger than the historical States would preserve the distinct cultural characters of their areas, while correcting the changes in scale that evolved as the Union grew and newer states became much larger than older ones. Each region would provide four Senators, giving an elite body of fifty-two members.  Senators would serve for up to three terms of 6 years each. Regions would be designated by agreement of the states encompassed by them, with anti-gerrymandering provisions that limit regional boundaries to existing state boundaries, or to new boundaries that must be straignt lines or prominent geographical features such as rivers and mountain divides, and with the limitation that the longest midline length of a region must be no more than six times as long as its shortest diameter. 

Eliminate the Electoral College 

Once both the House and Senate are representative of population and not land or historical accident, there is no need for the baroque structure of the Electoral College.  The President and Vice President shall be elected by the majority of eligible voters nationwide.  With 174 million registered voters, the likelihood of ties is so miniscule that elaborate tie-breaking procedures become unnecessary and would be eliminated as well.

The US is already well on the way to popular election of the President without waiting for a Constitutional amendment, via an Interstate Compact for a National Popular Vote.  US States can make agreements with each other without any need for Congressional legislation, although consent of Congress may be required for a compact to become finally effective. The Compact provides that each state allocate its Electors according to the outcome of the national total, which would make the Electoral Collage a redundant formality that could readily be abolished. It will take effect when it becaue agreed to by states representing a majority of the 270 Electoral College votes. In July 2026 it has obtained 85% of the necessary votes.  Each election in which the Electoral College becomes obvously unrepresentative of the national consensus brings the compact closer to implemenation.

Term Limits and Age Limits for All Federal Offices

The Constitution only specifies term and age limits for the President and Supreme Court, so it is not strictly necessary to have an Amendment to set these limits: they can be specified by law in the usual way. However, laws can be repealed or altered relatively easily; we would like to have limits that are more difficult to change.

Term limits should be defined to limit terms of service to 25 years or less, regardless of the length of individual terms.

Age limits should be defined based on objective criteria that apply to the population at large; being elected or appointed to public office should not constitute an exemption to the retirement rules that apply to the rest of us. Social Security and Medicare eligibility ages represent convenient benchmarks, yet no one should be considered instantly incapacitated upon achieving retirement age. Medical studies have found that mental decline typeically precedes death by 4-7 years, so if the population's life expectancy at retirement age (65) is about age 83, then the mandatory retirement age should be 77.  As medical progress continues, lifespans should correspondingly increase, so this age should be adjusted at 10 year intervals, in coordination with the decennial census results. 

Bonus: Open the Two-Party System with Preference Voting

Aside from the Twelfth Amenndment defining Electoral College procedures, and the Seventeenth Amendment providing for direct election of Senators rather than their appointment by state legislatures,  the Constitution is silent about how federal officials are nominated and elected.  Our two-party system of conventions and primaries is defined entirely by individual states and by the political parties themselves.

Many schemes to reduce the expanding extremisms that this system leads to have been proposed, and each has its advantages and disadvantages. But they all share the property of allowing voters to specify more than one candidate as tolerable, while still selecting a personal favorite.  When candidates can improve their own chances by advocating policies that may attract voters for whom they are not the first choice, the benefit of attacks on those voters' first choices is reduced.

Individual election jurisdictions can choose their own specific method of ranking candidates and selecting the most popular one. Every additional change to preference voting of some sort or another reduces the polarization of campaigns, and moves the country away from rule by extreme minorities, and towards the rule of the consensus that is the essence of democracy.

If the Constitution can't be fixed, then what?

If repairs to the Constitution don't happen, then there are three possibilities: the situation will improve, it will stay the same, or it will get worse.  Incremental improvement seems unlikely: the authors of the Constitution tried to build in defenses against the tendencies for politicians to overstep their power, but the damage that conservative Republicans have done to the good faith assumptions they made is probably irreparable. Second, the US will continue to muddle along as it has been doing.  Everyone will be unhappy, economic growth will decline or reverse, and the US will lose its dominant position in the roster of global countries.  Or, the Federal government will degenerate into a series of autocracies with a facade of elections, with laws established by a compliant Congress and courts that are suggestions rather than mandates, enforced depending on the latest whims of the bosses. 

After all, autocracies are simpler than democracies: people simply do whatever the big boss says, and they are free to do whatever they can get away with otherwise.  In a democracy, citizens have to consider the welfare of the other members of the nation.  This is considerably more difficult.  But it puts citizens in charge rather than bosses, and many of us prefer it that way.  Whether there are enough people who are willing to put out that extra effort so that "government of the people, by the people, and for the people" survives in the 50 United States remains to be seen. 

Do robotaxis increase traffic congestion? It's complicated.

In many urbanist discussions about the effects of the introduction of self-driving taxis into city centers, some commenter confidently pops up with the assertion that "robotaxis increase congestion".  Like most internet comments, this ignores much of the story.  They've probably read an article like this one, or less coherently this one [Isn't it strange how this site's "works in progress" never get taken any further towards final, definitive states?].

You could just as easily assert that before "robotaxis will increase congestion", there are "Uber and Lyft will increase congestion".  And then "personally driven vehicles increase congestion", too.  It's like "any kind of individual-ride transportation increases congestion".  The people making these arguments are conveniently forgetting about packed buses and subways, and have never seen the bicycle traffic jams in Dutch cities.

Perhaps this is all just a manifestation of Jevons' Paradox: any improvement in transportation capacity will increase demand by more than the capacity increases it provides.

Nor do these analysts take account of the fact that every ride not taken in a dedicated personal vehicle reduces the need for a parking spot at the destination. Although a taxi still needs its own parking lot during downtimes that lot doesn't need to be within walking distance of passenger destinations.  In a country where some cities have as much as 40% of their downtown real estate devoted to parking lots, this can make a big difference.  Riders will also appreciate that every ride in a vehicle that isn't dedicated to their sole use and doesn't need to be driven by the rider comes with automatic valet parking.  (Marketing of "remote parking assist" features make that claim somewhat nebulous.)

Induced demand ensures that as long as there are cities, there will be congestion. City governments and planners can mitigate congestion with street designs, mass transit, traffic light timing, and access taxes.  Robotaxis, and to a lesser extent human-driven taxis, add another dimension of manageability: because their dispatching and routing is controlled by an algorithm with access to the entire service area's data of availability and requests, the amount of congestion that those vehicles induce can be managed dynamically and in detail.  Individual drivers choosing when to begin their own trips can only be managed loosely, by statistical estimations based on their previous behavior.

Algorithms to manage taxi dispatching and routing can be as complex as their designers can imagine, but there are four basic varieties.

1. Dedicated vehicles. The rider has exclusive access to the vehicle. It waits somewhere nearby until it's needed, and goes wherever its rider wants, with no delay.  Any congestion that it may cause is out of the control of city officials.
2. Earliest available. Rides are provided by a pool of shared vehicles. As a ride is completed, its vehicle is added to the end of a list of available ones.  New requests are assigned to the vehicle that is the first one in the list, regardless of how far away it is from the requesting rider.  The vehicle might need to make a long "deadhead" trip empty to get to its assigned rider's pickup point, adding to congestion all along its route.
3. Closest Previous Destination. The algorithm scans the locations of the pool of available vehicles, and assigns the requested trip to the vehicle whose destination is closest to the requested pickup point. This minimizes the added deadhead congestion.  Congestion can be reduced even more by selecting the next ride from the full pool of vehicles rather than just those that are available at the time of the request, at the cost of making the requester wait a bit longer. 
4. Quickest Previous Destination. If the dispatching system has access to traffic data of the kind that "smart cities" were promised to have, it can look at the speed of traffic on all potential routes for all possibly available vehicles, and assign the ride to the vehicle that can get to the rider the quickest. If the system's traffic data is up to date in real time, it can choose a vehicle and route that avoids congestion, minimizing non-revenue-generating deadhead time and street congestion, and minimizing rider inconvenient waiting time. A triple win! Depending on the details of how the fare-determination rules manages the tradeoff between fixed payment per ride and variable payment per trip length, an experienced human taxi driver using tools like Waze can do much the same thing, but the results will be unpredictable.

When I was a freshman in college many years ago, I was privileged to take a "preceptorial" seminar taught by a senior professor of physics, in what I later learned was the field of "operations research" whose subject is modeling and analyzing this sort of scenario. In those days, when electronic calculators were provided to students in "the calculator room" and computers were accessed by submitting a stack of cards through a window and receiving a printout the next day, with an inevitable job abort indicated by cryptic error message, analyzing these scenarios in any more detail could take months of tedious work, so we were limited to simply thinking about the problem and writing down our conclusions, as this note does.

Today, anyone can run multitudes of simulations on a notebook PC, or even on a phone, and get back results more quickly than the time it takes to code the program.  Soon, a Large Language Model Artificial Intelligence will do the coding for you, and your job will only be to judge whether the results are credible or are the results of "AI hallucination".  If your goal is advocacy, you may not even care about the difference.

 

6-7 megatrends for the next quarter century

The first quarter of the 21st century is over!  For people who lived in a time when the twenty-first century was "the future", this is a shocking realization.  Things have not changed as much this time as they did in the first quarter of the 20th century, which saw the commoditization of the telephone, the invention of the airplane, and a global "war to end all wars" among other profound lifestyle changes.

But there are adults alive today for whom any year with a "19" as its first two digits is nothing but history.  The recent 25 years are their experience, and the next 25 are what they, along with the rest of us, are now concerned with.  There's no need to review the latest quarter century specifically, but we can use our experience to recognize a few changes that are going to continue into the next quarter century.  Not in any particular order, but using numbering to count how many different things are going on.

1. Artificial Intelligence: Large language models, task specific deep learning, and robotics. 
• Big tech companies are spending hundreds of billions of dollars on the misguided notion that if Large Language Models like OpenAI's ChatGPT, Anthropic's Claude, Google's Gemini, and Microsoft's Copilot can carry on a conversation as intelligently as a bad politician who doesen't know or care if what he says to voters is true or accurate, they can become sane and superintelligent by simply giving them ten or 100 times more compute power. 
• Task specific deep learning is likely to become pervasive. Machine learning systems are already transforming areas such as drug discovery, radiological diagnosis, and geological exploration for oil and geothermal resources.
• In work contexts, repetitive tasks are ideal for robotic automation, yet the ungainly locomation of bipedal humanoids is superfluous in most situations. Even the Jetsons' Rosie the Robot maid had wheels instead of feet. The current generation of robots lack a sense of touch, as robotics pioneer Rodney Brooks has bas oberved. Humans with similar limitations are called "disabled".
2. The decimation of US government-sponsored research.  For the past 75 years, the US government has pursued a strategy of sponsoring basic and applied research, which has made this country the global leader in innovation. This strategy has now been abandoned. Efforts by Trump and DOGE to destroy the government research-sponsoring establishment have been thwarted by the courts and by law-abiding bureaucrats following Congressional mandates, and have limited the reductions to single digit percentages in many areas.  Nevertheless, the World Intellectual Property Organization now ranks the US last in its rankings of innovation growth. China was first, while The United States and Japan experienced the steepest drops, losing 9.7 and 7.2 percentage points respectively, followed by reductions in Germany, France, Italy, Canada, the Russian Federation, Brazil, and the United Kingdom. Since the 1950s, a talented and energetic young scientist had been able to expect to build a career on a continuous series of government grants -- no more: that promise has been permanently withdrawn.
3. Climate change and its driver, fossil fuel energy. 
• Farmers and countries with significant agricultural sectors that deny climate change wil experience avoidable declines in crop yields. Megadroughts of 1000 years ago that lasted 100 years may last indefinitely this time. Heat waves that kill tens of thousands of un-air conditioned people in Europe and South Asia will continue, while humid Persian Gulf locations will reach effective temperatures incompatible with human life.
• Photovoltaic electricity will continue to be the cheapest power the world has ever seen. It and wind power, both cheaper than fossil-fueled power, will overcome political obstacles as greed prevails over ideology, while in low income countries off-grid and microgrid photovoltaics will be deployed without waiting for government or big business to provide stable grid power.
4. Healthcare decline. As global lifespans improve, they increase the number of people surviving to old age, when they will need assistance in daily activities.  Yet those same countries with longer-lived citizens also experience declining birth rates, with fewer young people around to take care of them. This will create a gap in elder care availability until household robots become advanced enough to replace care providers. There will be a difficult social transition as workers displaced by AI language models and other automation discover that care worker jobs remain abundant. The US abandonment of evidence-based public health policies will hopefully be reversed in future administrations.  Meanwhile, alliances of state public health departments have arisen to set their own policies independently of the federal government.
5. Autonomous vehicles. In 2025 the ability of automobiles to safely operate on city streets without human drivers was incontrovertibly demonstrated.  This milestone launches a technology transtion that may be as rapid as the 15-year demise of the human operated elevator, tempered by a replacement cycle for existing passenger automobiles that exceeds 14 years.  People who drive their own cars will become as exceptional as people who ride their own horses. Logistics companies have been complaining of shortages of truck drivers for decades; autonomous trucks will solve that problem, to the dismay of teamsters' unions which have relied on shortages to support their demands for higher wages and better working conditions. Autonomous flying drones are transforming warfare in Ukraine and elsewhere, while self-piloting air taxis will at last fulfill the sci-fi dream of flying cars.
6. Satellite megaconstellations will enable global wireless communications. Just as rooftop solar power enables less developed areas to leapfrog unreliable centralized power grids, satellite megaconstellations will enable those areas to leapfrog centralized voice and data networks. Remote areas can perform a double leapfrog, skipping over both wired and cellular infrastructures. But the price of never being out of touch will be never being out of sight of government and commercial surveillance. 
7. Honorable mention: the decline of the dollar as the global reserve currency. For more than 80 years, the US dollar has been the currency of choice for international transactions, due to the size and liquidity of its markets, and the stability that comes from a Federal Reserve bank that is independent of politically-driven fluctuations. President Trump is attacking that independence with all the means at his disposal, including criminal investigations of the Chair of its Board of Governors. Analysts are concerned that the value of the dollar relative to other currencies declined nearly 10% in 2025, although they fail to incorporate a perspective that it's up nearly 20% in the past 20 years. Yet gold and other precious metals, traditionally a haven in times of trouble, have doubled or more in their dollar prices in 2025. If the stability of the dollar can no longer be trusted, candidates for an alternative reserve currency include the Euro, the Chines Renminbi, and the basket of currencies that form the World Bank's Special Drawing Rights. None of these seem significantly more attractive at this time; a multipolar situation where they each are used for different purposes seems likely. That would allow politicians to claim dominance over "what's important" regardless of global totals.

Because socioeconomic systems include creative human actors who can choose "none of the above" for their decisions, their future is fundamentally unpredictable. Mathematical chaos theory identifies the rate of growth of unpredictability with a "Lyapounov exponent" that specifies how rapidly fluctuations diverge from their past trajectories, but economic theories incorporating this phenomenon cannot be found. Journalists and other pundits can make short term predictions, but how long anyone can expect those predictions to stay valid remains a mystery.

Unpredictability arising bottom-up from amplified micro-fluctuations in intrinsically chaotic systems is complemented by top-down shocks in hierarchically organized systems, where the impacts of dictatorial leaders or unplannable technological breakthroughs such as usable room temperature superconductors, cannot be timed or scoped. We must simply be aware of the possibilties, and be prepared to adapt to them if they chance to arrive.

Carbon rewards are almost here

Carbon rewards are very much like weather commodity futures, but denominated in climate change reduction units instead of Heating Degree Days.

If you were alert when you read Kim Stanley Robinson's novel The Ministry for the Future (you have read it, haven't you? If not, it's an entertaining tour through all the ways climate change could be mitigated), you might have noticed a mention of "Chen's paper" in one of the scenes with the meeting of central bankers in San Francisco. It turns out that Chen is a real person, Dalton Chen, and that paper and related ones actually exist in the real world as well as in Robinson's fictional world.

Called the Global Carbon Reward, it's a proposal for a new cryptocurrency that in the novel acquires the name "carbonis", backed by carbon mitigations, with central banks providing official credibility and administrative support to manage the correspondence between specific actions and currency distributions around the world, using bonds with hundred-year maturities to drive long term actions. As the Wikipedia entry for it observes, "The theoretical background to the global carbon reward is presented in several publications and a thesis, but the policy has yet to be reviewed by policy institutions or government officials." The papers appeared in 2017 and 2018, and yet nothing has happened in the following seven years. 

There appear to be a number of reasons for this. The papers gloss over the decentralized and complex nature of foreign exchange relations.  It's convenient to think, especially if you're conspiracy-minded, or a novelist, that currencies are controlled by a small cabal of bankers who regularly meet in Basel or Zurich to determine the future of global money. In reality, each one of them is a prisoner of their own organization's self interest in a system with so many moving parts and non-transparent elements that they can't be enumerated in a self-consistent way. Their entire careers are based on building and maintaining a façade that they know what they are doing, and that they really are in charge.  It's only after great catastrophes that small groups can cause major changes to the system, as was done at conferences such as Bretton Woods in 1944.  Chen's plan for a global climate currency entails a major change to global finance, and imposing it from the top, as Chen proposes, will require a global financial catastrophe that is possible via novelistic magic, but does not seem imminent in the real world.

To surmount these limitations, a global climate reward system:

• must be self-implementing. It must be able to grow organically from small elements.
• must be self-funding: It must immediately reward its participants, and not require participants to defer rewards for decades, as Chen's plan for hundred-year bonds requires.

Fortunately, we already have a system that comes very close to these requirements: environmental commodity futures.  A commodity future was originally simply a contract to buy a commodity such as a 100 bushels of wheat at some time in the future, at a price agreed upon today.  A farmer could use futures contracts to reduce the risk of price fluctuations at harvest time, when bumper crops could depress prices, or political events could distort demand. Futures markets arose in which those contracts themselves could be bought and sold.  Over the years, futures have evolved to become very sophisticated, and now many kinds of commodities are traded, including money itself, and especially cryptocurrencies.  They can now be written to be settled in cash rather than by delivering the actual commodity, eliminating the unfortunate situation where a trader who neglected to exit his position before it expired could be left with a railroad car full of grain.

 

The environment has become a tradeable commodity with the establishment of cap and trade frameworks that take care of all the tedious, complicated, yet essential groundwork of standardizing and certifying environmental mitigation actions. The establishment of a cap itself is not particularly important in creating tradeability -- what's important is the existence of a broadly recognized way to relate specific mitigation actions to defined climate effects.

 

Once there's an official, agreed measure of climate change mitigation, it's a small step to create market agreements and infrastructures that permit small pieces of mitigations to be bought and sold.  The fact that different organizations have different costs and timelines to implement kinds and amounts of mitigations creates opportunity for hedging and arbitrage that naturally lead to potentially profitable financial instruments that work by trading different forecasts of costs and effects.  The profit motive causes those instruments to appear almost by magic.  It's very close to Chen's Climate Reward, but without any need for central banks or crypto "money".

In order to promote long term solutions, Chen proposed that the carbon currency take the form of bonds that would take a hundred years to mature, in contrast with common bonds that mature in a year or five or ten or thirty years.  Dividends from those bonds would provide the reward component of the Climate Reward system.  This effect is weakened in reality, because once there exist enough of them, a market will arise allowing the bonds themselves to be traded, making their effective duration their age at the date of the trade, and giving them effective maturities across the full range of their duration.  The face value maturity date simply sets an upper bound on the duration that a bond may deliver returns.

 

Climate futures could obtain the same effect when their underlying climate model is run at differing forecast durations.  Each forecast duration would establish its own tranche of futures that would be traded independently.  There even exists a class of commodity futures that last forever, with no defined expiration date, called perpetual futures. For perpetual futures, a mechanism called the funding rate adjusts and pays the return regularly, their counterpart to bond dividends, although without the stability of bonds.  Chen's proposal assumes that these forecasts extend at least 100 years into the future in order to establish a value for its climate bonds, yet most climate models only run through 2100. Today's climate futures can only be realistically valued at 75-year expirations.

 

To launch climate futures, the key step needed to be done is to establish a certifiable set of mappings from mitigations to climate effects at a defined range of future dates, and to establish an organization to manage updating those mappings on a regular basis.  To be credible, the mappings will need to be created and documented by a team of qualified climate experts.  It will need to define equivalences between different climate influences in addition to carbon dioxide such as methane, water vapor, sulfates, clouds, dust, arctic ice and other factors that make significant contributions to global climate, and express them in a single value such as ton of CO2 equivalents.  In the absence of a regulatory regime that assigns a price to carbon, these will then need to be related to the value of a defined basket of conventional commodities whose yields are sensitive to climate, in order to establish a monetary value for each contract in the event that its actual delivery eventuates.

 

Once that's done, individual futures contracts can be written using those mappings and traded over the counter.  It doesn't take an official market to create those contracts as long as individual participants have the personal contacts to find people to trade with on their own. Once a sufficient number of over the counter trades are regularly occurring, an established commodities market can step in, define standard contracts in their trading system, and announce the market for these trades to be open. Voila! No waiting for bankers, government ministers, or legislators to pass new enabling laws!

 

Chen envisioned the Carbon Reward as serving a de-risking function on climate mitigation actions, making them easier to finance. Like all commodity futures, a climate futures market will serve its natural risk reduction function, and no one needs to rely on novelistic magic to save the world. 

 

Can alternatives to capitalism avoid causing civilization to collapse?

A: maybe. Why? Because all the other economic systems are unworkable. 

[An expanded version of this note is available on Medium [gift link]]

If capitalism is destroying civilization, then maybe alternative economic systems might be better. Even if human nature is magically changed for the better, popular alternatives like socialism and communism, and lesser known systems like gift economies that have worked for pre-technological communities of a few scores of people, have fundamental problems when applied to a global population of billions. If civilization is to survive, then we must discover how to somehow make the self-interested basis of capitalism work for survival, not against it. To survive itself, sustainable capitalism must be more profitable than unsustainable capitalism.

 The big picture of civilization's trajectory is easy to get lost in the sturm und drang of everyday international conflict with world leaders continually threatening nuclear apocalypse. Nevertheless, with lifespans in developed countries exceeding three quarters of a century, many of us will live through several shorter cycles and should try to plan for the longer trends. Here's a briefest possible summary.  The details are incredibly complicated, and intrinsically unpredictable, both mathematically and psychologically.  But there are some key poins.

Uncontrolled capitalism destroys itself 

Uncontrolled markets become monopolies and destroy themselves.  Sustainable markets depend on participants agreeing to not be losers, and creating governments and government agencies that prevent runaway monopolization.

Uncontrolled, pure capitalism reduces all activity to monetary transactions, which are intrinsically zero-sum with equal amounts of winning and losing, because accounting requires that the books be balanced.  The sum of all the debits must equal the sum of all the credits.  

Uncontrolled capitalism produces a spectrum of capitalists and workers with inequality that inevitably grows greater over time, because of hypercompounding: compound growth means that more capital you have, the more capital gets added to that at every step.  Modern electronic capitalism provides zero-fee investment accounts that permit even the poorest people to become capitalists and multiply their investments via compound interest.  But the rich become richer faster: More capital means greater access to resources enabling complex, fancy stratgies with even greater rates of return. 

As a side effect, uncontrolled capitalism produces waste that inexorably accumulates to levels that impair productivity growth.

Pollution continues to increase costs of production until costs exceed profits. Without net profits, growth stops, and the system shrinks.  It can be an orderly shrinkage, or a catastrophic collapse. If ways cannot be found to reduce the amount of pollution relative to growth, a cycle of growth and collapse becomes established.  Because much pollution can't be remediated, each collapse is deeper than the last, and the peak of growth is lower.  The cycle of growth and collapse has a long term declining trend, whose bottom is hard to predict.

Alternatives to capitalism contradict themselves, and don't scale 

Comprehensive socialism, AKA Communism, is intractable at scales larger than a few hundred people. Due to the complexity of communications required to make collective decisions, communist societies evolve hierarchical organizations, which inevitably lead to authoritarian, centrally-managed economies. Centrally managed economies require exponential amounts of computation to determine optimal allocation of productive resources, and their hierarchical organizations necessarily discard knowledge needed to account for local variations in requirements as they attempt to fulfil global goals. Moore's Law predicts that exponential growth in algorithmic power may be able to overcome the exponential demands of larger organizational units, but it cannot address life's hidden complexities.  All socialist societies to date have failed; the ones that have not failed, e.g. China, incorporate major capitalist components that implements decentralized economic decision-making.  

Regardless of computational concerns, communism's most idealistic promise to do away with money and allocate resources "from each according to his abilities, to each according to his needs" explodes under the paradox of "My need is to have more than anyone else."  What gives anyone, or any resource allocation committee, the right, or even the ability, to determine what my needs are?

Just as true socialists want to do away with private property, there's a possible mode of capitalism that would do away with private property for nearly everyone.  The recent explosion in purchases of property by private capital could lead to a situation in which nobody owns their own home except the very wealthy, and everyone else is forced to be a renter. The same transition could occur for what is most peoples' second largest purchase, their cars. These trends would lead to a situation reminiscent of medieval feudalism, with a small number of corporate "landed gentry" exercising lordship over a large population of propertyless serfs.

Zero-sum extractive capitalism relies on contradictory assumptions. There are not enough resources to go around: anything that I obtain must be taken from someone else.  Yet, there are unlimited resources to be taken; otherwise there is no growth. As the reality of a bounded world takes hold and resources run out, the capitalists's goal is to be the Last Man Standing, holding out until the bitter end after all rivals have passed away.

The rise of ultra-cheap renewable energy has led to the possibility of a petroleum apocalypse marked by the end of burning fossil fuels for energy, leaving a much smaller amount of production to be used for chemical products such as plastics.  Exxon Mobil, for example has been steadfast in holding to fossil fuel production, evidently intending to use its vast size to outlast competitors as global demand for oil and gas inexorably shrinks.

Can greed overshadow human nature?

Economies have their foundations in material, cultural and psychological values. Those values emerge as near-unconscious preferences for different solutions when people encounter resource allocation problems, which can bias or even override rational analyses of the best ways to divide things up.  Not subject to the arithmetic rules of monetary accounting, they can encompass a far wider range of aspects for each transaction, yet still produce a conclusion of "better" or "worse" in the end.  People vary widely in their selection and weighting of those aspects, but they can be roughly categorized into "win-lose" (zero sum) or "win-win" (positive sum) outcomes.

 

Some studies of the dictator game have provided for antisocial punishment, in which the "dictator" pays to punish other participants, as long as their counterpart ends up in an even worse situation. Everyone ends up losing.  Researchers have found that as many as 19% of subjects from English Speaking, Protestant European, and Confucian cultures will prefer the lose-lose outcome, while in Orthodox/Ex-Communist and Arabic speaking cultures such as in Romania and Greece as many as 53% of subjects will choose this. In the United States, political trends for the past decade or so suggest that this fraction may be as much as 40% of voters for many issues.

The problem for a manager, or a board of directors choosing a CEO, is when strategic decisions must be made, should they choose the path that feels good, or the path that makes them the most money? Politicians have an easier choice: they need to support policies that make their own supporters feel good.   Mercenary politicians soliciting campaign contributions need to convince mercenary contributors of the wealth-enhancing effects of their contributions, while using them to stoke the often-irrational passions of voters.

 

The drives for dominance are embedded incredibly deeply into Western culture, and pervasively mask the effects of cooperative growth and wealth. Even in supposedly objective economics textbooks, the fundamental principle that “people respond to incentives” somehow does not apply in the same way to workers as it does to executives.CEOs are given enormous quantities of company shares, while individual contributors are expected to be content with cost of living increases in salaries and wages.

Co-ops align incentives across stakeholders

Alignment of employee incentives with company performance reaches a zenith for employee-owned co-ops. In that structure, worker incentives are fully aligned with corporate success. However, there remain other contributors to corporate success whose incentives remain unaligned, and bringing them into the picture requires other kinds of organizations that are even more rare. Customer co-ops such as mutual insurance companies, or REI outdoor equipment, strive to put the interests of customers first. Supplier co-ops for farm production are not uncommon in the US Midwest, where one can see the Co-Op logo on grain elevators. Management groups routinely take over their own companies, although this usually occurs via leveraged buy-outs, making the true interest that of the holders of the debt rather than the managers responsible for sustaining the company.

To achieve complete alignment, a company would need an ownership structure that balances the interests of all five direct stakeholders: customers, employees, managers, suppliers, and pure investors. Achieving this kind of governance design is complex, and has not been attempted to my knowledge. A movement for “stakeholder capitalism” has achieved notable support, but it is still focused on voluntary and often token gestures towards inculding additional groups, rather than providing formal, effective control mechanisms for input into corporate decision making by groups whose actions can be explicitly stated and coordinated. 

Finance firms have a focus on total returns that can override irrationalities

Countering the irrationalities of “all for me” and “who cares who gets hurt?”, and the pseudo-rationalities of zero-sum no-growth economics, it finally may fall to financially sophisticated investment firms to realize the benefits of stakeholder alignment. They have the resources to perform the in-depth assessments to discover good management practices and share allocation policies. The most sophisticated of them will understand the nature of ergodic statistical distributions to detect the influence of long term perspectives in short term financial performance. Investment fund constructors and hedge fund managers who can recognize these advantages, and the stakeholders in their investment targets, will achieve growth advantages and become billionaires faster, and their clients and those who follow them will gain proportinately.

If capitalism cannot save itself, and non-capitalist forms of economic organization are unmanageable or unscaleable, a hybrid form may be able to moderate the worst excesses of each. Governments can establish regulations to stabilize runaway capitalist phenomena such as monopolies and the destruction of commons such as air and water quality, and fully take over sectors such as healthcare and common infrastructure like roads, water and sewer services and electric and gas distribution, and police and fire protection, where the conditions for functioning markets are not present. In 1787, the founding document of the current United States government stated that its purposes included “to promote the general welfare”. Many influential citizens and politicians appear to have abandoned this goal, but they do so at their peril. Is there a limit to how deeply civilization will fall once growth becomes impossible and collapse begins? No one knows the minimum number of people needed to support an advanced technological civilization such as the one that exists in developed countries. If we cannot make global economies sustainable at our current level, we will find out.

[2025-10-07 updated with a reference for antisocial punishment in the dictator game]

Prioritizing longevity research

 As one gets older, the effects of age begin to overtake the vigor of youth.  Those troubles are distracting and annoying, not to mention sometimes painful.  Inevitably, I suppose, they grow until the effort of sustaining daily life exceeds the amount of time and resources available to counteract and compensate for their penalties.  Ideally, a lifespan would be akin to that of the Deacon's Wonderful One-hoss Shay in the poem by Oliver Wendell Holmes, which operated perfectly for a hundred years, and then instantly and completely collapsed into a pile of rubble. The poem's author prefigured Star Trek's Vulcan philosophy in its final line "Logic is logic, that's all I say."  Makes sense to me, too.

Longevity researchers distinguish between lifespan and healthspan. The Holmesian ideal is that they are identical. Instead of ageing being a long slow decline, it should be possible to live a full, active life right up to the very end.  I've always imagined that being struck down by lightning while hiking in the mountains would be a satisfying way to go, except for the burden on rescuers who would have to retrieve my remains.

In short:

Longevity research shouldn't waste effort on making the longest possible life even longer. Nobody wants to spend decades of mental and physical disability, surviving in assisted living facilities on the efforts of ever more scarce younger, more able bodied caregivers as populations become older. (Robots will save us!)  This means that priority should be given to research aimed at chronic and disabling illnesses, affecting:

- cognition
- vision
- mobility, including metabolic declines in stamina and strength
- hearing
- immunity and organ replacement
- not to mention obesity per se

This by no means excludes basic research.  It's important to remember that the line of research that led to the GLP-1 receptor agonists that have revolutionized weight loss, was pivotally influenced by the discovery of a similar protein in the venom of the Gila Monster lizard, that had been incidentally described as part of a general catalog of the venom's components.

.

Why is simplicity so hard?

I used bound notebooks for note taking for many years, and accumulated a dozen volumes of them.  But I've moved online for everything except critical, mostly financial notes.  I pay for Evernote.  It's gratifying that they regularly find and fix bugs, but they keep adding features that I don't want and don't use.

Fred Brooks, in the classic book The Mythical Man-Month, observed that there comes a point in the lifecycle of large software systems where an attempt to fix a problem introduces as many or more problems as it remedies. Many of the advances in software engineering technologies are intended to raise the levels of scale and complexity that can be achieved before this threshold is reached.

A similar phenomenon has been observed in evolutionary biology, where it's called Cope's Rule.  Over time, members of lineages tend to increase in body size, until they get so large that they can't survive.  And of course evolution by random variation and natural selection leads to distributions of individuals that include members that are more and more complex over time, because while ecological niches may fill up, "there's always room at the top" for more complexity.

Evernote isn't yet at the point of net zero returns on improvement efforts, but they're working on it.  It takes a lot of management discipline to stifle the urge to keep adding "just one more cool thing" and say that a product is complete.  "Complete" is incredibly easy to interpret as "done for", i.e. dying.

My online note taking needs are simple:

Goes without saying:

• Notes must be searchable, both within and across notes
• Must have a simple editor that supports headings, indentation, bulleted and numbered lists.  That is, formatted with a WISIWYG frontend on a text backend

 Not so obvious:

• persistent - exportable to filesystem, e.g. in HTML
• collaborative - Syncs character by character across users
• sync among multiple platforms in use by a single user is a side effect
• structured - Three+ levels of nesting
• provides for links between notes and hyperlinks outside of the note-taking system
• multi-platform - Mac, iPad, iPhone, Windows.  Linux & Android optional

Wants:

•  versioned - "save a checkpoint now"

I'm tracking eighteen diferent note-taking apps, but haven't yet identified my next generation tool.  I have four months before Evernote autorenews for another year's subscription.

 

A strategic growth strategy for HP Inc. is boxed in on all sides

It's a big box, but it isn't growing.  And HP is too big to create new kinds of boxes, so the one path out isn't promising for them .

Journalists, tech pundits and investment analysts always thought of Hewlett-Packard as "PC and printer company HP", while it was far more than that.  So eleven years ago it split into HP Inc.(HPQ), an actual PC and printer company, and Hewlett Packard Enterprise (HPE), a company that did all sorts of computer-related things for other companies, including building supercomputers. 

HPE then dropped off the radar of everyone except specialist IT observers.  I wrote about its strategic limitations here in 2016, and those limitations haven't changed.  Its latest attempt to expand beyond its own box is an attempt to purchase Juniper Networks, which has been blocked by the US Dept. of Justice on antitrust grounds.

HPQ has two centers of focus, unsurprisingly desktop and notebook PCs, and printing on letter-size paper. Many people might be surprised to learn that its commercial aspects of these centers dwarf the consumer aspects, by three or four to one.  HPQ was recently in the news when it instituted, and then quickly dropped, a minimum 15-minute wait for free telephone support in five European countries. Although it triggered a lot of news coverage, the move didn't impact the bulk of HP's customer base, who provide their own customer support, and use HP's support services only for paid escalation incidents.

Strategically, attempting to economize on customer support expenses is a result of the box that HP is trapped in. PCs are a commodity product with fierce competition that keeps profit margins low. Top line growth is out of HP's hands, closely tracking economic booms and recessions. The only place available to increase margins is to cut costs, and expensive live support calls are an obvious target.

The commodity nature of HP's product lines, in concert with its great size, place severe limitations on its ability to innovate. Modern HP has never had visionary leaders to match its ancestral founders Bill Hewlett and David Packard, or even Compaq founder Rod Canion, who was fired in 1991. Its low margins and lack of vision mean that it doesn't have the time or patience to grow new businesses from tiny beginnings.  It has attempted to innovate by acquisition, but its history has been filled with inexplicable acquisitions of failing companies that it was unable to turn around.  HP's sheer size is a major impediment to innovation, since its organizational structure has been optimized to the point that any product line with less than several billion dollars in sales will be wiped out in the internal competition for resources.

Expanding HP's PC box with technical innovations is blocked in every direction: software, bigger and better screens, products smaller than the desktop, head-mounted virtual reality and augmented reality headsets, externalized "cave" virtual reality, and voice interaction".  HP has no room to lead in any direction.

Expanding the printing box with technical innovations has had more success, but nearly all of these directions have been exploited to their limits.  HP has no room to pioneer in the dimensions of bigger pages, faster printing, or 3D printing.

[You can find a version of this post with a bit more detail on these two boxes in an article at Medium.]

One frontier in printing remains open. Printing always occurs onto flat surfaces. Getting text and imagery onto non-flat surfaces, called "conformal printing", usually is done by printing onto a flat sheet, then applying that sheet to the target surface. Decals and vehicle wraps are widespread examples of this method. Methods without a carrier sheet that is applied to the object involve floating the ink on a liquid surface and lifting the object through the pattern, or coating the object with a photosensitive layer, projecting the image onto the object, and then chemically or optically fixing the latent image to the object.

Conformal printing techniques that build upon HP's proprietary technology could include mounting an inkjet printhead on a robotic arm that tracks the crests and dips of the object's surface, creating a flexible inkjet bar that sweeps across the curves of an object's surface, or exploiting a multi-material 3D printing system to create embedded imagery akin to the age-old millefiori glassworking technique.  Once multi-material 3D printing is available, enabling it to create embedded text and images requires only software development. Embedded text is especially suited to purposes such as serial numbers, since because they extend deep into the object, they cannot be removed without destroying it, although they can be covered up.

Conformal printing is a relatively small expansion of the market for flat printing, but it leads to enormous potential in its ultimate expression as active, smart paint, or more generally, smart surfaces, accomplished by extensions of each of those same three general methods.  The goal of those R&D programs is to duplicate and exceed the complexity, speed and dynamic pattern changes of squid & cuttlefish skin, and the adaptive textures of octopus skin.  Active smart surfaces promise uncountable opportunities that encompass capabilities that are unimaginable today.

Yet in the near term, the best available "strategies" that occupy the imaginations of investors are simply to track global economic and technological trends.  The latest standards in "AI PCs" promise to bring the unprecedented computational demands of machine learning to the desktop and laptop, while ML will enable self-optimizing printers.  These trends should drive a replacement cycle that can recover momentum lost to longer product lifespans of devices that have reached a maturity stage where customers' devices have become able to fully match the best that they can imagine.

Disclosure

I have held HPQ stock since HP Inc. was created by the division of Hewlett-Packard, reinvesting dividends as part of a buy-and-hold strategy.

The illusion of growth: "inflation" is only a third of the real story

 The Social Security Administration notified me of my next year's benefits a few days ago.  The notification letter asserts that my benefits increased by 2.5% over last year, but my take-home payment will actually increase only 2.07% after all the Medicare deductions.  Luckily, I invested my 401(k) retirement plan funds in index funds, so they are capturing some of the excessive profits skimmed off by insurance companies, pharmacy benefit managers, and other healthcare companies. 

According to the Bureau of Economic Analysis, "From the same month [in 2023], the PCE price index for January [2024] increased 2.4 percent. Prices for services increased 3.9 percent and prices for goods decreased 0.5 percent. Food prices increased 1.4 percent and energy prices decreased 4.9 percent. Excluding food and energy, the PCE price index increased 2.8 percent from one year ago."  So that 2.5% Social Security benefit increase doesn't seem totally unreasonable, aside from the fact that it's nearly a year out of date with respect to my current income and expenses.

So why do I, and so many other Americans, feel that our money isn't going as far as it used to?  Maybe because it isn't?  Unlike economists and economic journalists, we're concerned with value for our dollars, not merely prices, and not the rate of price growth that's called "inflation". And our memory extends longer than a year.

We remember this: since before the Covid pandemic in about 2018, the official PCE price index has risen over 20 percent.  And that index excludes gasoline and food, which have risen even more.  The PCE has risen over 40% since the most recent recession, and the amount of value a dollar provides has shrunk correspondingly.  I don't know who to feel more sorry for, the Americans who have to suffer from their loss of buying power, or the deluded economists and the clueless journalists who listen to them and tell us that our own experiences are wrong, and everything is good since wages have gone up too.  They don't care to notice that it's not the wages of the typical worker which have gone up, but the income from the wealth of the rich who have gotten richer far faster than the rest of us. And to add insult to injury, they promise that prices will never go back down as long as they have anything to do with it. The fabled rising tide doesn't lift leaky boats.

Loss of buying power is worse that rising prices suggest.  Lots of people have recognized "shrinkflation" when the same size package contains less stuff.  Marketers hope that you won't notice when the same size box creates an illusion that you're getting the same value for your purchase, while their profits go up because they give you less of what they're supposed to be selling.

Measuring shrinkflation is tricky, because it requires going into the details of how the data that composes an inflation index is compiled and entered into the index.  Sometimes an item may be measured by its container price, and sometimes it's measured by the unit price.  Some indexes may log a box of cereal by its price per box, and other indexes may log the cost of the same box of cereal by its price per ounce, regardless of the number of ounces in the box.  But even the most careful research organization is unlikely to count the number of chocolate chips per cookie, or the number and thickness of pepperoni slices in a pizza. 

The US Bureau of Labor Statistics has studied how shrinkflation affects the Consumer Price Index, and concluded that the effect is miniscule. When they looked at datasets that corrected for items whose quantity had changed between 2015 and 2019, they found the biggest effect in the category of Household Commodities and Services, which changed from 14.31 to 14.26, a difference of 0.05, about a third of one percent.  Considering how outraged consumer reporter articles are about shrinkflation, they are wildly overreacting.  I was surprised, too. It's possible that shrinkflation has increased along with monetary inflation since 2019.  Everyone should encourage the BLS to update their study regularly. In any case, it is annoying to say the least to discover that a manufacturer has pulled such a sneaky trick on you and effectively raised their price, hoping you won't notice. The fact that only a small number of manufacturers do such things doesn't make the practice any less offensive when it occurs.

The third form of inflation is even more insidious. Manufacturers can use lower quality ingredients or components in their products, and service providers can intentionally become sloppy or slow or reducing the feature richness in their work, while keeping prices the same or even merely lowering prices more slowly than they reduce their costs, thus increasing their profits.  There's no popular name for this practice, but "qualflation" works for me.  Some tech-oriented people might think of "enshittification", which was chosen as the 2023 word of the year by the American Dialect Society, but this word encompasses all forms of inflation. Whatever you call it, it's difficult for economists to measure, since it requires deep dives into supply chains and defining quality test procedures for every single one of the thousands of items that go into inflation indexes. Qualflation, because it's even sneakier than shrinkflation, is that much more offensive when it's discovered.  If you're outraged by it, your outrage is justified.

The remedy for all three forms of inflation is a free market with multiple suppliers producing a variety of variations on every service and product, or better yet, advanced markets with ecosystems of independent standards and unbiased testing organizations that produce regular reports. But a national currency doesn't exist in a free market -- there are commonly no choices of what money to use when buying something, except in rare situations such as the duty-free shops in international airports. And when governments allow monopolies or near-monopolies to arise, shrinkflation, qualflation, or simple price rises inevitably follow.

For complicated reasons, an economy with multiple currencies doesn't work well.  The United States had many private and state currencies until 1863, and many problems along with them. And since the release of a Bitcoin implementation in 2009, cryptocurrencies have been promoted as a new alternative to government-backed money.  Cryptocurrencies are working about as well for buying things as private currencies did in the past, that is not very well. "Crypto" is evolving into just another collectible asset of little intrinsic value, like fine art or Hummel statuettes. Like other collectibles, crypto has a robust free market, but Grandma isn't paying for groceries with Bitcoins any more than she is with Hummel certificates.

In the Federal Reserve Act (as amended), Congress gave the Federal Reserve System its fabled "dual mandate" of price stability and maximum sustainable employment.  The members of the Fed's Open Market Committee, in their wisdom, have decided to say that money with an inflation rate of 2% counts as "stable prices", and that anything lower is incompatible with maximum employment.  They prefer to try to fool everyone into thinking that growth is occurring, and will continue to occur, so that it makes sense to hire more people to cope with that future growth. Pay no attention to the slow loss of value in the dollar that results.  So far, they've succeeded, and consumer complaints about how high prices never go down have been successfully ignored.

Improving the best grilled cheese sandwich

In 2022, I wrote an article on Medium about The Two Secrets to the Best Grilled Cheese Sandwich. The two secrets are to do these two simple things: (1) use mayonnaise instead of butter, and (2) use real American cheese.  I included a bonus secret: use plain white, square sandwich bread.

That's about it. It's near perfection.  But "near perfection" still leaves room for improvement, or at least variation without loss of quality.  Here are two more things you can do to add variety to the perfect grilled cheese.

Using two slices of cheese, of course, sprinkle some crumbled feta or blue cheese or gorgonzola (which is more green than blue, to my eyes) between the slices before assembling the sandwich.  This will add a bit of salt and umami flavors, although American cheese is already pretty salty.

Try Fontina cheese instead of American.  This melts just about as well, but with a bit different flavor profile.  You could also try Mexican Chihuahua cheese, but you may have to go to a Mexican supermercado to get it.  It's a rare find in a mainstream supermarket, even in the Southwest US, where they often have an entire section of Mexican cheeses.

When something is so near perfect, almost every change is a downgrade.  These tweaks aren't downgrades, which is something else.