In the census of horses conducted by the Board of Trade in 1920, there were 19,743 thoroughbreds in Britain, out of a total of 2,192,165 horses. More than two-thirds of all these were draught horses, of which 774, 934 were in active agricultural work, ploughing the bulk of the 4.5 million hectares under cultivation (about the same area as today). By 1939, the number of horses on British farms had declined to less than a million, and during the 1950s virtually all of them disappeared — displaced, some say, not so much by tractors as by tractors with front-loaders. Those of us brought up in the 1950s are the last to remember working Shires, horse-drawn coalmen and rag-and-bone men, not to mention details like the hessian feed bags attached ’round the muzzles of horses during lunch break so that nothing spilled onto the clean suburban streets.
Britain, being at the forefront of civilization, yielded to petrol hegemony quicker than most nations. Western Europe has followed suit, but even in the 1980s you could still see, for example, a leathery French peasant guiding his yoke of oxen every morning along a certain stretch of the Route Nationale 9, not far from Rodez. I like to think they delayed the construction of the Autoroute 9 until the poor fellow was dead and buried, and in a broader sense they did. In Eastern European countries such as Poland and Romania, plowing with horses is still common practice, though under attack from EU modernizers. In the United States, horse cultivation is flourishing in the boondocks, inspired by the commercial success of the Amish communities, and aided by the fact that there is plenty of land to keep them on. In large parts of the Third World, draft animal power remains the only economic choice for small farmers.
It is at this stage in the evolution of farming technology that we are discovering that fossil fuels are causing more problems than they solve, and we are faced with need to find alternatives. “There are three horses in the race to replace petroleum – biofuels, electricity and hydrogen – and at various times you see the fortunes of these various horses ebb and flow,” a motor industry expert called Roland Hwang was quoted as saying in the New York Times. But however true that may be of the automobile industry, it will be some time before we see electric or hydrogen-powered tractors rivalling the use of diesel on the farm. The two main contenders are biofuels and the runner Hwang thought had been retired from the race – namely the horse.
These two sources of on-farm renewable energy are not incompatible, and there is no reason why they shouldn’t be carried out equally satisfactorily on adjacent farms. However, it is useful to compare their performance. The main problem is that one is faced with a superabundance of evolving data about the performance of biofuels, and a dearth of information about the performance of horses.
Draft Animal Power

Much of the casual commentary on plowing with horses is coloured by allusions to the amount of land they take up. We are variously told that when horses tilled all our land they required a quarter or a third of it to feed them. Darlington suggests that 40 per cent of farmland is required to feed horses, but this is derived from a secondhand reference to a conference held in 1973 in Alberta, where yields of crops per acre are extremely low. The most extreme example of this approach comes from one of the Global Opponents of Organic Farming, Dennis Avery of the Hudson Institute, who (citing Vaclav Smil) states:
“If American farming were horse-powered today, it would take 250 million horses to match our tractor and harvester engines. We’d need 740 million acres of land to feed the horses – twice as much arable land as the United States has. Instead of exporting food to densely populated Asian countries, the United States would be hard put to feed itself.”
Although this statistic is meant to illustrate the inefficiency of horses, it doesn’t take too much thought to see that it actually reflects the inefficiency of U.S. agricultural machinery and its extravagant fossil fuel expenditure. 250 million horses is almost one horse per U.S. citizen, enough to plough 2.5 billion acres, which is enough land to feed the world. If the U.S. is employing that much horsepower to produce its crops, it needs a radical rethink.
The most reliable calculation probably comes from Jan Jansen’s study of the agricultural economy of a Swedish village in three years during the 20th century. In 1927 there was one working horse for every 8.4 hectares of land, and horses consumed 18 per cent of all the energy harvested in the community. However, there is little to be gained from estimates of the amount of farmland devoted to horses in days gone by because crop yields have increased dramatically, while the area which a horse can cultivate has probably increased somewhat as well, thanks to improved technology. In 1920, one and a half times as much land was put down to oats as to wheat (whereas now the oat crop is negligible) and this was a reflection of the need to feed horses. But in 1920 the average yield of oats was 1.7 tonnes per hectare. Now the UK average is 4 tonnes for organic oats and 6.5 tonnes for nonorganic, and it is even higher in Ireland. Leaving aside the hay and grass that will always be a proportion of a horse’s feed, we now need roughly a third as much arable land to feed a horse as we did in the 1920s.
So how much land does a working horse need nowadays, and how does that compare with a biofuel tractor? It is difficult to calculate because it depends how much oats and how much non-arable grass it is getting. Ken Laing advises: “A starting point would be to allow 1 tonne per horse per year for a horse worked frequently” with the rest of the feed consisting of pasture or hay. One benchmark for comparison is to take a fairly high yielding field of wheat and determine how much horsepower and biofuel power it could, theoretically, generate. Wheat yields in the UK are around 8 tonnes per hectare, against 6.5 for oats (which is what horses prefer); but since horses also eat grass from nonarable areas, and since the feed value of oat straw is more than of wheat straw, the figure of 8 tonnes is reasonable. Eight tonnes of grain, grown on a hectare of prime land, will provide 22 kilos of grain every day for a year, which is enough in terms of calories (72,000 per day) to keep two horses working at a moderate pace.
Two horses are normally agreed to be capable of cultivating ten hectares. According to horseman Charlie Pinney, “when farm horse numbers were at their highest, the generally accepted horse-per-hectare ratio was around one pair per ten hectare, with that number increasing by one horse per each ten hectare increment in farm size. If the average farm unit is 40 hectares, it therefore needs five horses,” which is 16 hectares per pair. It looks as though, theoretically speaking, it requires the equivalent of one hectare of good quality agricultural land to provide the horse power to cultivate 10 hectares. This figure is confirmed by Jansen who calculates that his Swedish village, were it entirely horse-powered today, would need to devote 10.2 percent of its arable land to feeding its horse, plus another 0.8 for the energy necessary to manufacture the equipment.
Biofuel Farming
How does this compare with biofuels? This is even trickier to determine, not from a paucity of information, but because there are so many different kinds of biofuel, and so many different analyses of their potential performance. A hectare of land, producing about 8 tonnes of grain, according to figures from what was then called the Department of Trade and Industry, which are the most optimistic I can find, will produce bioethanol equivalent to 1,000 litres of diesel. It takes about 100 litres of fuel to cultivate a hectare of wheat, and if we allow another 10 litres for the embodied energy cost of the machinery that means a hectare of bioethanol could power a tractor to cultivate that one hectare and a further 9 hectares.
Further Considerations for Plowing With Horses
There is therefore a broad equivalence between the energy performance of plowing with horses, and the energy performance of running a tractor on bioethanol from wheat. Circumstantial factors and mitigating circumstances could be argued on both sides. Horse advocates could point out that nearly all studies view bioethanol processing to be even less efficient than the DTI figure I have used, while biofuel advocates might point to studies showing that if the wheat straw were used to provide heat, or distillers grains’ anaerobically digested, then the process would be more efficient. Horse advocates might argue that the cultivations necessary to grow ten hectares of grain shouldn’t take a pair of horses more than 100 days, leaving another 160 days (plus two rest days a week) on which the horses can do moderately heavy work at no extra fuel cost – whereas any additional work performed by the tractor would require extra fuel. Tractor advocates may riposte (illogically in this context) that you don’t have to feed a tractor when it’s in the shed.
And so it can go on with, I suspect, the balance in favour of draft animal farming at the moment, simply because it is an existing and proven technology, whereas on-farm biofuel is not. However, vastly more investment and research is being directed into biofuels than into animal technologies, and it is anticipated that within the next few years a “second generation” of biofuels will outstrip the performance of existing methods. Whereas current biofuels are generated inefficiently from high value crops such as wheat, corn, rape oil and (less inefficiently) sugar cane, the second generation will focus on converting cellulose from fibrous crops such as switchgrass and Miscanthus, which can be grown on lower-quality land. Protagonists claim that Miscanthus can “produce about 2.5 times the amount of ethanol we can produce per acre of corn.” If that is the case, then there is a high chance that it will prove to be more efficient than draft animal power. Jansen examines a range of biofuel options for his Swedish village, which require between 14 percent and 6 percent of the arable land to produce the energy needed for manufacturing and powering the machinery. The low figure of 6 percent is achieved by using willow coppice rather than rape oil to power the machinery. The fact that a horse emits about a fifth as much methane as a dairy cow might also weigh against animal traction.
Concluding Points on Biofuel Farming vs. Draft Animal Power
If the intrinsic efficiency of biofuel farming over draft animal power is confirmed, then it still remains to be shown that the tapping of this energy does not require an infrastructure so energy intensive that it defeats the object. A significant advantage of draft animal power is that it requires only the animals themselves, a few steel tools that last for generations, and a bit of leather – and animals reproduce of their own accord. In a study of horse-powered farming systems, Chet Kendell found that on a30-acre farm in Michigan, over a period of 40 years, a farmer derived a revenue of $21,000 from the sale of his horses’ progeny, whereas a fossil-fuel-powered farm of the same size, whose small tractor is traded in for a new one every 10 years, would have incurred costs of $70,000. On-farm processing of biofuels for a tractor would require additional equipment, though the expense of this would be offset by savings on fuel bills. Centralized generation involving the transport of biomass to power stations, and the subsequent redistribution of the energy and other products back into the wider economy, however great the economies of scale, might have a hard time matching the innate sustainability of on-farm biofuel systems which, like working animals, produce and expend the energy on site.
The contest between draft animal power and biofuel farming is a fascinating and important one, but it is bizarre how the cards are stacked. Animal traction is the dominant way of cultivating land throughout about half the world, while biofuels are barely used at all, except in Brazil, and experimentally in countries like the United States and Germany. Yet hardly any mainstream research is carried out into the comparative efficiency and sustainability of animal traction, whereas virtually every university in the western world has a faculty investigating the potential of biofuels. There are hundreds, if not thousands, of academic papers comparing different kinds of biofuels, but, aside from Jansen’s study, I have yet to find a single one analyzing the efficiency of draft animal power, which, after firewood, is the most frequently used bio-energy in the modern world. In such a prejudiced climate it seems inevitable that biofuels will emerge the winner.
The late Charlie Pinney, who for many years was a rare voice in the UK propagating the benefits of horse drawn cultivation, observed:
“The widespread nostalgic appeal of the draught horse often proves to be the biggest hindrance to the popular acceptance of its serious worth as a farm, forest and transport tool. Whenever you approach an individual or a government to suggest that the working horse can make a worthwhile contribution to a sustainable future for us all, you are almost invariably greeted with scorn or disbelief.”
The reluctance to examine draft animal power may also reflect a worry, on the part of scientists, that their efforts to turn biomass into energy through diverse highfaluting systems will prove no more efficient than the digestive systems of biological animals. The resistance of scientists is reinforced by the opposition of economists who view livestock as labour intensive: It takes a skilled man to handle a team of horses cultivating an acre or two a day, whereas the average farmworker these days drives a 140-horsepower tractor capable of covering 20 times the area.
The social ecologist, however, will view the matter differently. Time may be money, but speed, though lucrative, is disproportionately expensive on energy; and as regards energy, it is land that is in short supply, not humans. More people on the land means less people on the streets; and people on land produce energy, whereas people on streets consume it.

In his groundbreaking book “Meat: A Benign Extravagance,” author Simon Fairlie explores whether the hypothesis that vegetarianism is better for human health and the environment holds true. Based on numbers from the UK, Fairlie’s well-founded scientific research explores the difficult environmental impact of eating meat along with the ethical and social issues surrounding the future of farming livestock across the globe.
This excerpt has been reprinted with permission from Meat: A Benign Extravagance, published by Chelsea Green, 2010.
Something went wrong loading the comments for this article, we're looking into it.