Many naturally occurring phenomena follow a normal distribution and are therefore well described by a “bell curve”. Since the average length of men in Sweden is 180 cm, a high proportion of Swedish men are in the range of 175-185 cm. Some are shorter or longer, but most are in any case between 170-190 cm. If we broaden the span further to 165-195 cm, not many men at all are outside of that range. The same principle of course also applies to women and children (e.g. “weight curves” for newborn babies) and many other characteristics where there is a random natural variation.
The road network of Sweden is in a similar manner more or less “normally distributed”. On a map of Sweden’s road network (below), most medium-sized cities have a handful of roads leading to and from the city. Some have only three, more cities have four or five roads and a few cities are nodes that have six or more roads leading to/from them. The number of roads is thus fairly evenly distributed and there is only a difference of a factor of two between having three roads (which are few) and having six roads (which are many).
The “opposite” of a normal distribution is a “power law” distribution. With a power law distribution, it is no longer meaningful to try to locate an average and the difference between “few” and “many” will differ with a factor much greater than two. If we look at the network tying together Europe’s airways, we see a very different pattern compared to the road network, with a small number of very well-connected airline hubs and a large number of “feeder lines” that connect smaller cities/airports to a single or a few hubs. Large hubs in the European airspace – Frankfurt, Amsterdam, London and Madrid – link together a large number of destinations near and far. In Sweden, Arlanda airport outside of Stockholm is just about the only hub where people transfer between international and domestic routes.
Just as the normal distribution appears all around us, so does power laws. Every year there are innumerable small earthquakes, hundreds of medium-sized, but on average only one major earthquake (more than 8 on the Richter scale). Sales of cultural “perishables” such as published books, music, computer games and box office (cinema) tickets also follow a power law distribution. A small number of computer games (books, movies, artists) account for the majority of all computer games sold, and of all copies of a specific computer game that are sold, a disproportionate number are sold the first weeks and the first months – after which the numbers taper off into a trickle.
The business literature does not talk about power laws, but about the 80-20-rule; 20% of all X account for 80% of all Y:
- 20% of those who call customer service account for 80% of all complaints
– 20% of all salespersons account for 80% of all sales
- 20% of all criminals account for 80% of all crime
- 20% of the population own 80% of all resources
– 20% of all software bugs in a computer program account for 80% of the problems
– 20% of all websites account for 80% of all traffic
It is moreover in the nature of power laws that 20% of 20% of all X account for 80% of 80% of all Y – that 4% of landowners own 64% of all land and 4% of all software bugs account for 64% of all problems. The bottom line is that the differences between “much” and “little” can be huge – even if it doesn’t always follow that the distribution is precisely 80-20 – sometimes it can just as well be 70-30 – but it is primarily the principle of uneven distribution that is of interest here.
The structures of these two types of networks (the highway network as described by the normal distribution and the air network as described by power laws) are thus very different. Which brings us to the concept of peak oil, centralization and localization. In former times, slaughterhouses, bakeries, breweries and dairies were small, numerous and more or less evenly distributed across the country (just like roads are). Now they are big and they are located to only to a few places. According to the Swedish National Food Administration there exist (only) 25 “large-scale slaughterhouses” in Sweden that they oversee (and some of these slaughterhouses are – in line with the power laws – very much larger than others).
Six of these 25 Swedish slaughterhouses are run by Scan, one of northern Europe’s largest food companies (working mainly with meat). As if by coincidence, I read in a Swedish trade journal (“Agriculture News”) that one of Scan’s six facilities (in Uppsala) will be closed and that most of the activities at another facility (in Skara) will disappear in the near future. “The slaughter and butchering of cattle and sheep [will] be centralized to Linköping and the slaughtering and butchering of pigs will mainly be located to Kristianstad. […] Total headcount will be reduced from 3 000 to around 2 500. Some of the employees may be offered jobs at one of the locations where activities will be centralized.” In a column in the same issue of Agriculture News, the journalist Erik Brink writes appreciatively about Scan’s plans for restructuring their business:
”closure of all slaughter in Skara is just the next step in Scan’s crusade to get the Swedish meat producers to better adapt to market conditions. All old emotional trash will be cleared out and only that which makes business sense will remain.”
“Adapt to market conditions” is in this context equivalent to going big and “large-scale”. If you follow the transports going to and from these (currently) 25 large-scale slaughterhouses you will see that they are hubs with many long transports going to and from them (similar to the air network). There has been a push to rationalize and scale up the size of plants for decades, always moving towards an equilibrium that basically depends on the relationship between the cost of energy and the cost of labor. Energy has been cheap and labor has been expensive. In the search for higher profits, the tune has been to scale up operations by rationalizing, streamlining, concentrating, consolidating and slashing jobs – despite the fact that increased concentration also leads to increased vulnerability and longer transports. As energy has been cheap and the supply has been stable for decades, there has been little reason for any afterthoughts… until now.
Because what will happen if energy becomes more expensive in the future? What if the relationship between labor and energy will change radically (albeit gradually) after peak oil? I do not mean that gasoline prices (or the price of electricity) will go up with 10 – 20 – 50 – 100 percent, but that the relationship between labor and energy will change fundamentally. Human ecologist Folke Günther has made a graph that shows how many seconds a Swedish industrial worker has had to work in order to buy a kilowatt-hour of energy in the form of gasoline at the gas pump:
Two interesting things become apparent in this picture. The first is that “we” (an average Swedish worker) have had to work less than a minute to buy a kilowatt-hour of energy in the form of gasoline since the 1950’s. The second is that the relationship between work and energy changed by a factor of 10 in less than fifty years (between 1920 and 1970) and subsequently has remained at a low and stable level since then. In parallel with this drop in absolute prices, the use of fossil fuels increased by 1000% (a factor of 10) in Sweden during just the second half of these fifty years (from 1945 to 1970) when the welfare state was built. Energy Engineer Bengt Randers has collected the equivalent information regarding the cost of energy in the form of electricity and has a data set that goes back all the way to the 1890s:
Without violating his figures too much, it is possible to state that the cost of one kilowatt-hour of electricity in terms of time spent working in order to buy it was slashed by half in just one decade from the mid-1890s (from two hours to one hour). And the cost was halved again over the next 10 years (30 minutes in the mid-1910s). And the cost was halved again over the next 10 years (15 minutes in the mid-1920s). And the cost was halved again over the next 10 years (8 minutes in the mid-1930s). And the cost was halved again over the next 10 years (4 minutes in the mid-1940s). And the cost was halved again over the next 10 years (2 minutes in the mid-1950s). And the cost was halved again over the next 10 years (less than 1 minute in the mid-1960s).
From one kilowatt-hour of electricity “costing” 120 minutes of work in the 1890s, the same amount of electricity has during the last 45 years cost less than a minute of work – a difference of more than a factor of 100! In a picture showing the same trend on a logarithmic scale it is easier to see how the fast drop in prices ended in the 1960’s and how the price has remained exceedingly low since then.
How much then is a kilowatt-hour of energy? A man working hard physically and taxing his muscles can generate 75 Watts (Pimentel and Pimentel, “Food, Energy and Society“, 1979) and thus has to provide almost 13 hours of hard physical labor (no breaks) in order to generate a kilowatt-hour of energy. To exchange less than a minute of work for a kilowatt-hour of energy has thus been a very good deal (thank you oil!). To say that energy has been free for half a century is only a marginal exaggeration. In a previously published blog text [not yet translated to English], I refer to studies showing that a tractor can do in one hour what it took 19 days of agricultural/physical work (150 to 225 hours?) to do 200 years ago.
During the last fifty years, the cost (in time) of energy (gasoline and electricity) has thus been both extremely low and stable – but that will no longer be the case after the effects of peak oil assert themselves. It is probably correct to say that “everything” will change after peak oil. The energy-consuming hubs that I have described above (airports, slaughterhouses etc.) will become difficult to maintain if energy becomes too expensive. They will then shrink in size and be replaced by many smaller local businesses that are built on a business model that requires less energy and less transportation. In an ideal world, we would support such a shift already today:
”[We should give] support to small, local organic farms where the labor is largely provided by humans and animals, products are marketed to nearby communities, the plants and animals are raised in diverse polycultures that deter pests and preventable diseases, the animals feed the soil with their composted wastes, the soil feeds the plants, and the plants in turn feed the animals in a tight recycling of wastes and nutrients.”
If energy and transportation becomes 2, 5 or 10 times more expensive compared to the costs of labor, then much more work would be done locally solely because of economic pressures in that direction. (A piece of advice for those thinking about post-peak oil careers: in addition to local food production, local small-scale food processing that enhances the value of the raw agricultural produce will become a growth industry in the future!)
Just as rationalization has mechanized agriculture and depopulated rural areas, so will more labor intensive (food) production methods exert pressure in the opposite direction – away from (especially the largest) cities and towards a more evenly dispersed population living in smaller towns and in the countryside. The population will thus not be so concentrated to a few big cities (power laws), but will be more evenly spread out across the country (normally distributed). If just a fraction of tasks now done by machines will once again be performed by humans, this will require many more people to live in rural areas in order to produce the food we need.
Up until this point in time, human labor has gradually been replaced by highly efficient machinery, large-scale facilities and long-distance transportation. An underlying driving force has been that people/labor is expensive because energy (and hence production and machinery and capital) has been cheap. In each situation where it has been possible to save money by cutting the workforce, it has been rational to replace yet another person with yet another (energy-consuming and capital-binding) machine. This is obviously true not just for big machinery in industry and agriculture, but also in the service sector. As each and every person has become a “knowledge worker” equipped with a computer, that person has also to a greater extent been forced to handle his/her own administration – because all secretaries have been fired.
Where rationalization for decades has been equivalent to labor rationalization, we will in the future be forced to rationalize based on entirely different criteria. Labor rationalization will be replaced by energy rationalization in all activities and at all levels. While this text is called “the death of rationalization”, the title actually refers only to a specific kind of rationalization – labor rationalization. Where labor was previously rationalized away and replaced by energy-wasting habits, the pendulum will start to swing in the other direction when the costs of energy starts to rise.
When the consequences of peak oil start to make themselves known, it will become cheaper to hire an extra person if that reduces the energy consumption or the capital tied up in machinery. If not sooner, then it will eventually no longer be profitable to replace a machine that (finally) breaks down with a new machine instead of having an extra person at hand. If I try to think of some oil-powered “luxury machines” that can easily be replaced by strong arms and legs, I for example think of lawn mowers. At Husqvarna’s website, a hand-powered lawn mower costs 135 USD, a motorized lawn mower costs between 400 and 800 USD and a lawnmower that you sit in and drive costs between 3400 and 8800 USD (prices and models are from the Swedish market/homepage and may differ elsewhere).
A perhaps even more challenged gadget is the leaf blower that Americans in particular have a fondness for. It is a (noisy) gasoline-powered “backpack” that acts as an inverted vacuum cleaner that blows away leaves and debris from the sidewalk. At some point in time it will be more sensible to hire an additional (flexible, teachable, versatile) person and “invest” in an extra rake or broom instead of buying a new leaf blower (that needs to be refueled, serviced and repaired regularly). Moreover, it is already now possible to question the efficiency of leaf blowers (see text “Grandmother proves rake and broom as fast as leaf blowers”).
Some tasks, on the margin, will no longer be sensible to perform at all. At what point will it no longer be possible to direct an armada of vehicles to plow away newly fallen snow from all streets and roads in Sweden? And let’s not even talk about gasoline-powered toys such as jet skis, snowmobiles and off-road motorcycles…
With a shift in the labor/energy equation, other tasks (professions) will be added to the labor pool. I think you can get a glimpse of the future of Sweden/Europe if you look for all the tasks that are (still) performed by people in less affluent countries (elevator boy, doorman, concierge, guardian of parked cars) – but that in Sweden nowadays are performed by a machine (dishwasher, washing machine, vacuum cleaner, car wash) or by yourself (polishing your shoes, sewing up a pair of trousers, cleaning, cooking).
The author Agatha Christie have summarized both my text above and developments throughout the 20th century when she wrote in her autobiography that “When I was young I never expected to be so poor that I could not afford a servant, or so rich that I could afford a motor car“.
One blog asked what will correspond to Christie’s car and servant 100 years from now. Some took for granted that business-as-usual will continue and that energy and production will become even cheaper relative to labor in the future:
”I never thought I would be wealthy enough to own a holo-deck [or “a private jet”], nor so poor I could not pay someone to cut my hair [or “to go to a restaurant”].”
Others (fewer) drew the same conclusions I draw – that the direction will change, that work (services) will become cheaper in relation to gadgets, and that we will see more servants and fewer cars in the future. But there are also those who believe that everyone will own a robot servants of their own in the future … I do not think we will have any robot servant at all, because there will be an abundance of human servants to choose from in the future – just as when Agatha Christie was young. But since all of these predictions (including my own musings above) are based on more-or-less intelligent guesses about the future, my favorite prediction was this one:
”I won’t be rich enough to own a varnox. But I will won’t be so poor that I can not afford a glip-thorp.”
This text is dedicated to Oscar Kjellberg who has thought about issues relating to energy, economics and labor for a long time.
PS. From one of the largest daily Swedish newspapers, November 17, 2009: “Arla builds record dairy”
The dairy group Arla will build the world’s largest dairy outside of London. Arla Foods UK’s dairy will be four times larger than the largest dairies in Denmark and is planned to have a capacity of one billion liters of milk a year, the group informs. The plans form part of Arla’s strategy to significantly expand UK operations. The dairy will be finished by 2012 and will then have about 500 employees “