Sunday, February 23, 2014

Growth is a Ponzi Scheme

A persistent theme in conservative literature is the idea that unending population growth is essential to keep the economy and work force growing. A typical example is this by Don Feder in "Feder's Rules For Conservatives (Part I)" (GrassTopsUSA November 29, 2012):
All issues are connected. Abortion isn’t just a social issue or even a moral question. It undercuts the family, which devastates the economy. It also eliminates future workers and consumers, reducing growth.
It's hard to know where to begin untangling the ironies here. We could start with the practice of seeing human beings, supposedly too precious to abort, being valued solely as cogs in an economic machine. But there's a far more colossal irony. It's a trivially obvious mathematical fact that nothing can grow indefinitely in a finite environment, and especially nothing can grow exponentially for very long. The numbers get ridiculous very fast. If you could double a sheet of paper 42 times, the resulting stack would reach to the Moon. If everyone made 10 copies of a chain letter and sent them to friends who did the same, and it took a week per cycle, in a few years the amount of paper would fill the observable universe. So the huge irony is that people who deride Social Security as a Ponzi Scheme insist that the recipe for continued prosperity is the ultimate Ponzi scheme.

A Ponzi scheme is one where a con man convinces people to invest, and pays them off from the income from new investors. Eventually, the recruitment slackens or investors want to cash out, and the scheme crashes. Bernie Madoff probably didn't set out to run a Ponzi scheme. More than likely, he thought he was really clever enough to out-perform the market. When things went bad, he probably believed the old mojo would soon return. Like a system gambler in Vegas, he thought he was brilliant when he had merely been lucky. On the other hand, there are plenty of Ponzi schemes that are run by outright frauds who count on being able to hide overseas, shielded by bribable governments. Oh, one way to increase your own take is to pose as an investor and pay yourself first.

It's not a Ponzi Scheme if you incur debts in the reasonable expectation of paying them off on legitimate future income:
  • Borrowing money for a mortgage is not a Ponzi scheme. It gets borderline Ponzi if you buy a house way beyond your means in the expectation that your salary will increase or the value of the house will skyrocket. But even so, there's a reasonable expectation of actual income. It's risky, but not Ponzi. What makes it emphatically not Ponzi is that sooner or later you will pay off the mortgage. Unless it goes underwater, and you walk away. In which case you're merely a thief.
  • It's not a Ponzi scheme to borrow money to expand a business in the hopes that additional income will allow you to pay back the loan. Risky, but not Ponzi.
  • It wasn't a Ponzi scheme for the Catholic Church to expect that young nuns could be used to help care for elderly ones. It worked for centuries as long as the flow of new recruits continued, and if women entered convents at the same rate, they could go on forever. The system collapsed when women began to have other alternatives.
  • Social Security is not a Ponzi scheme. As originally planned, it would be pre-funded, but conservatives, the forerunners of the ones calling it a Ponzi scheme today, balked at the government accumulating such a huge surplus. They insisted on it being a pay as you go system. Right now, the system is stressed by the influx of Baby boomers into the system. But there are sensible proposals to deal with that issue. They are mostly being blocked by people who would like to crash the system. Their plan essentially is to steal the Social Security fund to pay off their own taxes. The only way Social Security can go broke is for American voters to elect people who let it.
It is a Ponzi Scheme if you plan on indefinite future growth:
  • Financial Ponzi schemes either promise steady growth or unrealistically high rates of growth. Someone who promises to double your money in a short time obviously has to double the size of his investment fund. It might be a good idea to ask how he plans to do that.
  • Unlimited population growth is a Ponzi scheme. All those people will want places to stay, food to eat, and jobs, which population growth advocates say will come from future growth. It's ironic that conservatives resist the idea that population might have limits, since every single threat to their liberties stems directly from population growth: crowding, crime, congestion, demand for services and regulations. The reason Wyoming or Alaska are so different from California is entirely due to low population density. If libertarians want the freedom we had in 1880, we'll have to go back to the population density we had in 1880. At this point someone throws in the Straw Man, why don't you volunteer to leave? Who are we going to kill off? Answer: nobody. Reduced birth rates will do it nicely, and that's happening in most developed nations.
  • Pyramid business schemes, where participants are rewarded for recruiting new members, might be legally different from Ponzi schemes (it is possible to profit if you get in early and recruit enough members) but morally they are absolutely indistinguishable. Sooner or later the last level of salesmen down the line will be unable to get anyone else to sign up and will be stuck.
  • In the long run, every economic model that assumes unlimited future growth is a Ponzi scheme.
Apart from the unrealistic promise of perpetual growth, most Ponzi schemes have another feature: a phase of deceitful maneuvering designed to conceal the inevitable collapse as long as possible. Bernie Madoff seems to have been quite aware for a long time that he was basically running a Ponzi scheme. He got away with it by mailing out fictitious statements to investors and paying off the relatively few who decided to cash out.

Growth in Nature

Put a few goats on an isolated island. This has actually happened with results that have environmentalists tearing their hair out. The goats eat the vegetation, and with lots of food and no predators, they have lots of kids. They in turn have more, and the population increases faster and faster. This is positive feedback; every change results in yet more change. The growth is exponential. But eventually the goats consume all the available food. The goats are weaker, less able to bear healthy young, more vulnerable to sickness. The baby goats are weaker and more likely to die before adulthood. Growth slows. Now every new goat makes it harder for the rest to survive. This is negative feedback; every change tends to inhibit further change. The growth rate levels off and begins to decline. The environmental factors that obstruct growth are called environmental resistance.

At this point several things can happen. It may be that the goats destroy so much vegetation that they all starve. More likely, the last few litters of goats in the exponential growth phase push the population beyond the level that the island can sustain and there is a massive but incomplete die-off. The population crashes, the vegetation recovers, the population grows again, and it may repeatedly spike and crash. In the most favorable outcome, population growth gradually slows and reaches a limit. A graph of population over time has a steep initial phase of exponential growth, a more or less linear period where the growth rate slows and then declines, and a period of steadily slowing growth. The graph of population over time looks like a slanted S, and a growth pattern of this sort is called a sigmoid or logistic growth curve.

That's what's happening to human population in the developed world. People had more and more kids (back to the human kind, again) to help on the farm, who in turn had kids of their own, and population grew exponentially. But soon population began to encounter resistance. Land became so fragmented it was no longer feasible to farm. Even in the Middle Ages, laws were passed limiting land inheritance to the eldest son. And the others? They went to the towns, which were population sinks. Anyone who looked at parish birth and death records could see, and did, that deaths outnumbered births in the towns. Obviously we have a lot more people now, but that's a topic for the next section.

Even with better food and medical care, there are limits to family size. Living quarters get more crowded (making it less practical to have children). Children cost more to feed, and the very growth of adult wages means children cannot really contribute to family income any more. Instead of contributing to family income, children drain it. So people begin having fewer children. The growth curve flattens. The point where population growth stops accelerating is called the Demographic Transition. It's ironic that good old fashioned materialism may just save the environment by putting the brakes on population growth.

Exponential growth, like perfectly elastic collisions, or bodies sliding without friction, or falling without air resistance, does not exist in the real world. It's a mathematical idealization that is useful for modeling some aspects of reality. On our imaginary island, some goats won't have young and some kids won't survive. A graph of goats versus time will be approximately exponential, but not exactly. The decay of a radioactive isotope, which is an example of exponential decay (the exponent on the exponential function is negative), is very close to exponential when there are billions of atoms, but won't be nearly as exact when it comes to the last hundred.

Growth curves won't be exactly sigmoid either, but sigmoid curves will be far more realistic than exponential growth in one all important respect: they remain within finite bounds. True exponential growth doesn't, and the numbers get very ridiculous very fast. An old story tells of a king who wanted to reward a loyal follower, so he asked what the follower wanted. Instead of the princess or half the kingdom, the follower merely asked for a grain of wheat on the first square of a chessboard, two on the next, then four, and so on. The king thought this was a trivial request until the court astrologer did the math and discovered there isn't enough wheat in the world to cover all 64 squares. At which point the king probably had the loyal follower's head chopped off.

But Wait!

I hear you cry. That may be true of goats on an island, but what about compound interest, which does follow exponential growth exactly? Well, the growth does, unless the borrower declares bankruptcy. But more than that, interest cannot grow forever. Imagine you invested one cent in 1 AD at 1% a year and cashed out in 2000 (to keep the numbers round). Your penny would be worth 4.4 million dollars. Well, okay, some extremely rare coins sell for that. But the average Roman coin is worth only a few dollars. Let's say your penny is invested at 2% per year. By 2000 it would be worth $1.6 quadrillion, far more than the GDP of the planet. That would be about 105 trillion ounces of gold, or about 3.6 million tons, far more than has ever been mined. You couldn't demand payment in gold because there isn't that much gold in the world.

So mathematicians have recognized, pretty much since compound interest appeared, that it can't grow forever. Mostly this is an abstraction. Your bank account interest grows until you withdraw the money. Your loan accrues interest until it's paid off. Nobody leaves money in the bank for 2000 years, because no bank has ever lasted 2000 years. But beyond the short term finite nature of most compound interest are larger scale interruptions like default or devaluation. I have an interesting array of valueless hyperinflated currency like a German 2 million mark note from 1923, Greek drachmas from 1944 overprinted with new denominations, a 50 billion dinar note from soon-to-be no more Yugoslavia, etc. When the state that issued it ceases to exist, or the currency is abolished, so does the value of the currency.

Real Growth

The reason we have a lot more people living a lot better than they did in 1400 is we have had centuries of real growth. We've colonized new continents, developed new devices and agricultural methods, and so on. How much growth has there been? Most estimates give a world population of 300 to 400 million in 1400, compared with the roughly seven billion now, so global population has increased roughly 20 fold in 600 years, or about 0.5% per year. 

What about economically? This is harder to pin down, because even the poorest countries have things like cell phones, electricity and automobiles that didn't exist centuries ago. We also have 20 times more people. British economist Angus Maddison estimated the GDP of Western Europe to be about $11 billion in 1100, quadrupling to $44 billion by 1500. The present GDP of Europe is about $18 trillion, roughly a 1600 fold increase, or a bit over 1.2% per year. So with a 20-fold population increase, that means roughly an 80-fold per-capita increase in GDP. Compare the US ($48,000) with Liberia ($600). Not hard to imagine our forebears in 1400 having a Liberia-level lifestyle. An 80-fold increase in living standards in 600 years works out to about 0.7% per year.

So we've had six centuries or so of GDP growth at about 1.2% a year. But wait? What about all those fortunes made by traders and mill operators? Those were big gains that directly affected only a few people. If 10 per cent of the population sees a 10% rise in their wealth, that works out to 1% averaged over the entire population. Still, there is some truth to the trickle down theory and the idea that a rising tide lifts all boats, especially once entrepreneurs figured out there was more money to be made in selling to mass markets than catering to the elite.

Prolonging Growth

Some of the earliest innovations in capitalism revolved around shipping. That's where we get expressions like "my ship came in" and "second rate." (The term referred to ratings of seaworthiness by insurers, and second rate was actually not bad. As opposed to sixth rate, which might possibly stay afloat for a while when untied from the dock). You invested in a voyage. If it came back, you got paid off. End of story. You spent the profits or went looking for another voyage to invest in.

But suppose you're running a factory that makes widgets. There's an initial phase of steady (roughly exponential) growth, which is great fun. But eventually, people have all the widgets they need. You sell just enough to replace the ones that wear out and your growth flattens out. The sigmoid growth curve again. So what's to stop you from just being satisfied with the steady income from selling all the widgets people need? It might work. If you pay back all the investors (buy them out). If some upstart doesn't come up with a way of making better and cheaper widgets. If widgets don't go out of fashion, or get replaced by something else. And if you have investors, because you need their money to cushion emergencies and modernize your plant, they'll want steady dividends and they'll want to see the values of their shares increase. If not, well, there's another ship sailing somewhere, and they'll jump ship and climb aboard. In an economy where new ventures are starting all the time, investors can grow their wealth by moving from one to the next. So like it or not, every business in our present economy has to find some way to grow. Kodak, a company that did incalculable good in its day by allowing us to record our memories and our history, is close to extinction because it failed to realize until too late that digital photography would completely demolish its business model. Kodak might have survived indefinitely if film had remained the primary medium for photography. (Although I just bought a Kodak printer that impresses me very much. They may just make it. I hope. All together, now, "Mama don't ta-a-ake my Kodachrome awa...y")

If you can't continue to grow a single business, diversification is one fairly benign way to keep growing. You make sausages, so you make more and more kinds. And you branch out into other meats. And you start making machinery for processing meat. And you buy packing plants. Diversification not only allows growth but insulates against market variations. If bratwurst goes out of style, you fill the gap with other products. A bit less benign is conglomeration. One of the earliest examples of conglomeration was Standard Oil. Why "Standard?" Because kerosene, which is what the principal petroleum product was in the late 1800's, consists of molecules about twice as heavy as gasoline. The lighter the molecule, the more volatile. Too many lanterns and stoves were exploding because many refineries did a sloppy job of distilling off the light molecules. Between the variable composition of petroleum from different oil fields and the sloppy distillation, it was pretty haphazard. John D. Rockefeller realized that to produce a reliable product, he had to control the entire production chain from well to distributor. One of the first big monopolies was created by a robber baron to create a safer product. (It wasn't totally altruistic, of course, because when people realized his kerosene wouldn't blow up, they'd flock to buy it.) Ain't reality a b****? This was conglomeration of related firms, but many recent examples involve companies buying up completely unrelated firms simply to have a place to invest profits. Thus, R.J. Reynolds, which made tobacco products, merged with Nabisco, which made cookies, to become RJR Nabisco. Then, ironically, the tobacco branch was spun off again into a separate company to protect RJR Nabisco from liability for tobacco settlements.

Pseudo-Growth: When Real Growth Falters

Eventually the growth slackens. It's at that point that business practices start to take on Ponzi traits, resorting to gimmicks to forestall the realization that the ride is about to end. One option for keeping things going might be called churning (a term that's actually used in the financial business). Churning in the stock market consists of making lots of unnecessary transactions that don't benefit the client much but do generate commissions for the broker. Cell phone and internet providers convince customers to switch plans (and in some cases switch them without their consent) simply to generate new business. Planned Obsolescence is a time-honored variety of churning. All products have finite lifetimes, but those lifetimes could be extended substantially by better manufacturing, availability of replacement parts, and so on. The profit from selling a car with an expected lifetime of 20 years, including replacement parts, doesn't equal the profit from selling two cars for ten years each. The cost of building a rustproof car of stainless steel might be twice the cost of a standard car but the impact on profits from having cars that lasted for decades would be far greater. A homemaker is much more likely to buy a new set of china if it's impossible to buy single pieces of the old style.

Software development has pretty much reached the end of innovation and is mostly churning. Why did Microsoft try to replace the perfectly functional Windows XP with Vista? And then with the okay but only marginally better Windows 7? And then try to replace Windows 7 with Windows 8? None are significantly better at the user level than XP, and the real improvements in functionality have been internal and visible only to software and security specialists. So why weren't they simply released as upgrades? Because Microsoft has long since reached the end of real growth. If they simply released upgrades, they'd be limited to selling occasional new copies of XP and releasing updates. They'd have a handful of employees and a fraction of their revenue. The only way to maintain the illusion of continued growth is to pretend minor upgrades are really new products. And most software has reached this point. Once you have a text editor that can do any typographic task, what's left to do? Once a spreadsheet can handle any calculation, what new features are there to add? Most software is plagued by "feature creep," the continued addition of new features that may appeal to a few users, solely to avoid the obvious conclusion that software evolution for established products is pretty much stagnant.

You can also try to take over part of an already viable market by stealing market share. All those school kids who come to your door on fund raising ventures are merely trying to cut into the market pioneered by the Girl Scouts, or they're selling products already sold by local dealers, like pizza. They have nothing new or innovative to contribute, and indeed many are supplied by commercial firms whose sole business plan is to try to take market share from already existing retailers. The most flagrant cases these days involve software. Google is about as efficient as a search engine can get. The only thing it could do to improve would be to stop collecting, storing, and sharing search results. So what do Bing and have to offer as rivals? Absolutely nothing. They exist solely to try to grab some of Google's market, and with it, advertising dollars. Even worse are some of the fly-by-night search bars that can infect your desktop. Many of these actually use Google to do the searching. Their sole purpose is to steal market share. Why does Apple feel the need to create a mapping feature when Google's is well developed and better? Because Apple needs to fend off admitting that it's stagnant.

One way to keep growing, at least a little bit, is to try to whip up demand. Advertising has degenerated into a frenetic scrabble for ever diminishing returns, as advertisers seek to target consumers ever more precisely in the hopes of snaring a few new ones. In the process, media degenerate from vehicles for entertainment and information to mere advertising outlets.

When Growth Stops

Sooner or later, the opportunities for real growth stop. What then? The first step, like with someone on a starvation diet, is to eliminate fat. So all those middle managers who were axed had to go. Then you consume muscle. Pension plans disappear, health benefits shrink, compulsory overtime increases instead of hiring new workers. You hire workers for 30 hours a week to avoid paying benefits. The good jobs are abolished when workers retire and the new jobs pay less with fewer benefits. All those cool, employee-friendly dot-coms that crashed around 2000, with the climbing walls and the employee day care and the flex-time were consuming muscle even in their prime.

Of course, the ultimate in consuming muscle is simply to butcher the cow and barbecue it. And lots of investment capital firms do just that. They buy up functioning companies, sell off the assets, kick the employees to the curb, and keep the profits instead of paying shareholders. Mitt Romney's old employers, Bain Capital, were infamous for these practices.

One strategy to forestall an end to growth is parasitism. A business can seek to keep revenue flowing by demanding continuing payments for what used to be one-time purchases. Recently some real estate developers have proposed that each house sale impose a fee to be paid to the developers, so that not only does the developer profit from the initial sale of the house, but from every sale thereafter. Performing artists have pushed for royalties from radio stations each time their works are performed (and not, as they properly should, from the record labels). Adobe's decision to make Photoshop a subscription-only product rather than a licensed product is a classic example of parasitism. Indeed, all software licensing is parasitism. If you owned your copy of Windows 97 outright and could sell it to someone else, why would anyone buy a new version?