Things that Can’t Go On Indefinitely Stop

Politicians have to act as though they know the solution to lowering unemployment and raising economic growth rates.

It is hard for anyone running for an election to admit that they might be powerless to affect some energy and economic realities. In this post, I discuss the trend in the figure below: US monthly personal-consumption expenditures (PCE) for food and energy goods and services as a percentage of total household expenditures. I think it is completely possible that the stop in the declining trend of PCE for food and energy that stopped in the early 2000s is indicative of the new reality facing the United States energy and overall economic (and debt) situation.


It doesn’t take a PhD in statistics (or engineering, or anything else) to notice the major change in the trends of the time series in the figure. In 1999, the 4-decade trend of decreasing PCE for energy stopped declining and started increasing.In 2007, the (at least) 45-year trend of decade trend of decreasing PCE for food stopped decreasing – just before the beginning of the Great Recession that began in late 2007.  Adding the PCE for “food + energy” shows a minimum PCE percentage of 11.7% in the first two months of 2002.It is a good question of whether or not this will be the minimum percentage PCE for “food + energy” for the US … for all time.

Several analyses have been done investigating a seeming threshold percentage of US PCE that can be spent on energy goods (and services) before it induces or plays a large role in causing a recession (see James Hamilton’s blog entries here (July 14, 20102)here (March 6, 2012)here (September 19, 2012), and there’s another one or two or more in there somewhere tracking the same things). But there is much value in considering the role of food in PCE along with energy PCE because food is both the original source of pre-industrial power and it holds high priority in a “hierarchy of needs” sense.


The reason to add the PCE for food and energy is because food is technically an energy source.It is too bad that the statistics needed to plot the data of the figure even further back than 1959 are not readily available, but it is practically certain that the percentage of PCE for food continues higher as one moves back further in time. Before fossil fuels and significant industrialization using wind, wood, and water power in the early 1800s, food was the major energy resource for prime movers. These prime movers were the muscles in humans and animals doing the majority of the physical (‘useful’) work and providing the most aggregate power.Thus, the quantity of food and fodder produced from the land had the major influence on the amount of power for agriculture and a little industry. The book Heat, Power, and Light: Revolutions in Energy Services, by Roger Fouquet, shows a similar historical graph for the United Kingdom in which he estimates the expenditures for domestic power in the 1500s as at or above 100% of GDP. Thus, in preindustrial times, practically all GDP was to produce power, or useful work!

The data plotted in the figure should be shown in the United States presidential election debates. I’d like to know what our leaders think of this graph and why they think their policies can or cannot affect the trend (or rather new trend since the early 2000s). The likely truth is that demographics and resource constraints have caught up with much of the ‘advanced’ economies (e.g. EU, US, Ja

pan). There are less young people working to pay for older people to retire.There are fewer older workers retiring because their pensions and retirement funds are not sufficient, thus not making room for new and younger workers. The conventional oil alternatives (oil sands, deepwater, oil shale, biofuels) don’t have the same level of pure energetic value of those of the past, and this is the reason that oil prices must remain at current levels in order for new oil supplies to remain viable. I wrote in a 2011 article in Sustainability (see here) how energy return on investment (EROI) and oil price are related. Alberta oil sands are practically the marginal oil supply (in North America at least) and because the oil sands have EROI < 4-5 then the oil price must be near or above 90 $2012/BBL. The economy remains sluggish, unemployment is still below 8%, and the jobs that people are getting are lower paying.

I think that both the ‘extreme’ left and right are wrong in their approaches. The U.S. can’t borrow money and go further into debt to give people high paying jobs while lowering employment, and the U.S. can’t borrow money to maintain the same defense budgets of either the Cold War or recent post-Cold War eras. Resource constraints are imposing their real nature upon our real economy. No lowering of interest rates can prevent this impact, and this is why the unprecedented length and level of low US and EU interest rates are not having the ‘expected’ effects. What we see is that demand for fuels and energy services have decreased (particularly less light duty vehicle miles traveled) since 2008 at a peak near 3 trillion miles (for a nice graph of US vehicle miles traveled see above link to Hamilton blog on Sept. 19, 2012 and search Stanford research on ‘peak travel’).

A ‘bottom’ percentage PCE on food and energy. A ‘bottom’ of interest rates.Coincidence? I think not.

Article reprinted, with permission and minor edits, from EnvironmentalResearchWeb.

Photo illustration, via USDA, is in the public domain.

The author, Dr. Carey King, researches energy systems and how they work together and within the environment.

His research interests focus upon:

(i) relating measures of net energy to economics;
(ii) understanding how technology and policy can interact within the nexus between energy and water;
(iii) integration and transition to increased renewable energy production;
(iv) the economics and life cycle of system-wide/integrated carbon capture and storage infrastructure; and
(v) promoting objective analyses of energy tradeoffs for energy education, decision-making, and policy development for natural resources.

Dr. King is currently a Research Associate at the Center for International Energy and Environmental Policy at the Jackson School of Geosciences. He works as part of several other collaborative research groups at the University of Texas at Austin: The Webber Energy Group ( and the Gulf Coast Carbon Center ( of the Bureau of Economic Geology.


  1. I am interested in this analysis, but would find it more useful if the impacts of higher prices were disaggregated to show the effects at different income levels. In increasingly stratified societies, what counts is not so much the total cost, but how many are excluded. Is this sort of analysis available anywhere?

  2. There's a lot more to this story. "Energy as the master resource" approaches seem a little too sure of their assertions to me. All interesting hypotheses, but I'm uncomfortable with them being stated as established fact.

    The story has a history also covered by Engel's law / the Engel curve and the general subject of "the proportionate distribution of expenditures under conditions of increasing or decreasing income” (Zimmerman 1932 p.80). The pattern being discussed covers a longer period shifting from a mainly agricultural economic base to one where it became a tiny slice of economic production (as well as diversifying into processing networks).

    The role of cheap energy is vital to that story but it's not the only factor. More important, maybe, is the rest of the economy continuing to develop: as more of total economic production is non-agricultural, so agriculture is relatively smaller - and the percentage of personal income spent on food, on average, thus has to be smaller.

    A couple of other statements I think are less settled than suggested:

    "There are fewer older workers retiring because their pensions and retirement funds are not sufficient, thus not making room for new and younger workers."

    There aren't a finite number of jobs. New workers should be able to feed into the development of new work.

    "The conventional oil alternatives (oil sands, deepwater, oil shale, biofuels) don’t have the same level of pure energetic value of those of the past, and this is the reason that oil prices must remain at current levels in order for new oil supplies to remain viable."

    Maybe, but the current level isn't all that historically high (here's a UK analysis of that). Fuel is going to remain *relatively* cheap for a good long while yet (long enough for climate disruption to overtake it as a serious concern).

    There needs to be a more systematic debate between different economic approaches. Eco-economics analyses come with a lot of axioms that worry me (many of them rather handily collected in one place in the CASSE book `enough is enough)' in the same way that anyone assuming peak oil is worrying - if they turn out to be wrong, it implies drastically different outcomes.

  3. I think if we were to stop regarding ourselves as the center of the universe (except to our personal selves, but that's a psychological rather than biological fact, I think) but rather as part of an integrated system with altering symbiotic properties, the whole thing makes a bit more sense. I believe living in harmony with our surroundings and each other is key.

    Exploitation is no longer working. It's kind of like the fable about the race where increasingly slow modes of transport finally advantage the slow and steady over the fast and lazy.

    Maybe it's just time to let go, but I am getting so fed up with the frantic efforts to prove that we will get a different result by running in place with illusion. IMNHSO, we don't need experts in taking, but experts in giving and sharing.

  4. I appreciate the comments on the blog that Tom posted for me regarding the percent of personal consumption expenditures (PCE) on energy and food. My main goal with this plot is to create discussion on what are the most important underlying processes at play in terms of the downward trends stopping in ~ 2001.

    To Peter T: I want to perform your suggested analysis. A student is beginning to look at that question.

    To Dan O:
    I don't think there are a finite number of jobs available, but there is a finite amount of time for any one person to work. If there is then a finite number of people on the planet, there is a finite amount of man-hours. Further, when the population eventually peaks (yes, I believe it will), there will by definition be a declining number of man-hours for work (which is OK because you'll need less goods and services).

    Also, the graph itself does not imply anything about energy being a master resource. However, energy is one of many resources that can be limiting. It takes energy to move things and the economy is really about moving stuff around, so the discussion of what comes first (energy vs. economy/population/etc.) is often a discussion held in the context of modern economies, and it should be in the context of a longer time frame. But I do agree with you (I'm guessing) that it takes energy efficiency to be able to afford more expensive resources. If I want to be able to afford $150/BBL oil, then my best chance is to drive a 50+ mpg vehicle and not a 25 mpg vehicle.

    Also, I do believe that if there were no fossil fuels, then the planet would not have 7+ billion homo sapiens on it right now.

  5. Thanks for participating, Carey.

    My question is about food prices. Yes, certainly a point will probably eventually be reached where there isn't enough food to go around, and that will surely drive prices up. Factors contirbuting to that crisis are: fossil fuel shortages for direct energy use in cultivation, fossil fuel use in the chemistry of fertilizer production, soil depletion, and climate disruption. Indeed, I am among those who argue that climate is a factor in the recent food price spikes that have destabilized the Middle East.

    But that is not the only factor. A very severe inequality in distribution of financial resources causes people to compete for their own necessities against others purchasing luxuries. Anyone who frequently traverses the American plains in a car (as I do in my pilgrimages between Texas and Chicago/Madison) is aware that most of the agricultural production of the area, one of the world's most productive, is dedicated to feeding animals, not humans. Crop-fed meat is (in addition to being ethically dubious on several axes) a luxury, not a necessity. As a vast middle class emerges in China, demand for meat has greatly increased. So even though a unit of meat has ten times the crop demand as a unit of cereal, there are plenty of people willing to effectively consume ten times their share for their own preferences.

    Now, we have countries dependent on money-based foreign aid to feed their populations. Any glitch in production on earth (typically weather-related, but potentially politically caused as well) means that these countries essentially are competing on a fixed budget for things they need that other people simply want.

    If "money", whatever that is, were roughly evenly distributed, this would matter very little. But if middle class people are willing to spend more than ten times as much on meat they want as poor people are able to spend on grain the need, the logic of the market calls the want a higher use than the need.

    I guess this isn't news to anyone who follows the topic. But there are those who see present-day food scarcity as population-driven or resource-constrained. It isn't - if global per capita meat production has held steady, contemporary food production would be excessive and prices would be too low.

    My conclusion, though, is somewhat surprising. We have two choices to maintain stability and avoid starvation 1) we can declare food a human right and create a mechanism whereby everybody is fed 2) we can come up with a mechanism whereby every society has comparable wealth. Neither of these approaches is market driven.

    With neither of these in place, if I am a moderately wealthy person in a moderately wealthy society and you are a poor person in a poor society, the free market will conclude that my luxury is more important than your starvation. It will effectively hide your tragedy from me, and will fail to offer me any mechanisms to help you. The greater the income inequality, the sooner you will starve, no matter how much surplus capacity there is in food production.

    So to get back to Carey's graph, what we are seeing on the food axis may be very different than what we are seeing on the energy axis.

    On the other hand, the extent to which all commodity prices operate in lockstep is another question, which has had me scratching my head for a while.

  6. "Declare food a human right"
    How funny is this? We sure can't allow that, hahaha... -- Hm, second thought: Reproduction should then be also a human right. But what about reproduction when food is scarce? Example: During the Syrian drought before the ongoing civil war the birth rate was highest in those areas that received most food aid. Conclusion: You can have a human right on food or on reproduction - but not both. You get food aid only together with "voluntary" sterilization. (Is the human race that smart? I guess no - we prefer war and starvation over reason.)

    "Factors contirbuting to that crisis are: fossil fuel shortages ..."
    Fossil fuel shortage plays a double role. Here is the other one: Example: Egypt was able to finance and subsidize food imports until recently. That way the Egyptian population could grow way beyond the local carrying capacity. (My guess is tenfold.) And now they have a problem.

    How many people could we feed globally by non-destructive non-fossil agriculture? That is, in a locally non-genosuicidally sustainable way that leaves soil fertility for future generations, without dependency on dwindling resources like mined phosphorus or high density energy for ammonia synthesis. (BTW that would be a novelty in the history of mankind.) -- I'm not so sure the answer is easy.

  7. Conclusion: You can have a human right on food or on reproduction – but not both. You get food aid only together with “voluntary” sterilization.
    Odd conclusion. Mine would be 'areas where insecurity about the future, as measured by the need for food aid, have higher birth rates, so reduce population growth rates by improving stability and security'.

  8. Carey

    Thanks for responding. My question was sparked by putting this together with an observation by Brad deLong on the calories the wage of an English unskilled labourer could buy (not much more than enough to keep him alive in the C18; hundreds of times more than that now) with the analysis of Floud and Nagel in The Changing Body. If F&N are right, food production over most of Europe's history was enough to keep everyone fed, or some people well-fed if others lived in near starvation. Food was the fuel that kept society running up to the mid C19, and oil and coal do that now. So the equivalent question to "how much grain does a wage buy?" is "how much energy (oil or coal) does a wage buy? and how has that changed?"

  9. Pingback: EcoEquity: Global Climate Justice » Book Review: Countdown: Our last, best hope for a future on Earth?

Leave a Reply

Your email address will not be published.