I am presently in Morocco for research purposes. As the things were getting a bit slow in the last few days due to the public holiday meant to celebrate the birthday of the Prophet, I decided to indulge in my geekiness and I undertook a micro research. I am fascinated by the elegance of economic geography; I enjoy watching the market spread in space as much as others like to watch birds; so that’s just what I did, my Excel sheet in one hand and my binoculars in the other (for the birds, you never know).
I used 224 of the 227 cheeses proposed by the on-line French cheese-seller fromages.com to try to confirm empirically von Thünen’s theory on agrarian location. My hypothesis was the following: some cheeses are easier to transport than others, it is thus reasonable to expect that the most difficult to carry and conserve are produced close to their consuming market.
To test my hypothesis, I created a “transportability index” (I know it’s ugly).
The cheeses were rated from 0 to 3 according to their hardness. The cheeses with a soft paste and a soft crust got a 0, those with a soft paste but a hard crust or the Roquefort-like persillé cheeses got a 1, the uncooked “pressed pastes” were rated 2 and the long-lasting cooked pressed pastes” received a 3.
I then assumed that, considering the importance of Paris in the organization of the French market, the distance from the producer to Paris should be used as a variable. Unfortunately, the information provided by the website regarding the production location were pretty vague and only indicated the administrative region the cheese came from. The mainland is divided in 22 such regions. The list also included cheeses from Corsica, the Netherlands, and Switzerland, which were counted as regular regions. I used the regional capital city as a proxy for the location of the producing areas within the regions. I could not weight my data either due to the absence of production quantities. These shortcomings mean that the result given below should be regarded as general trends rather than very precise figures.
A first regression analysis including all the data proved disappointing 0.281 (t=8.222). I then assumed that the Parisian influence on the market may be limited within a certain range. In a radius of 600 km, the results are already much stronger: 0.493 (t=8.974). Within 500 km, they become very significant: 0.745 (t=8.960), but fall at 400 km: 0.638 (t= 8.554). Finally, at 300 km, the results reach 0.809 (t=4.998).
The French cheese production appears thus organized around Paris but only within a 500 km radius. The graph is particularly telling in this respect: the transportability of cheeses increases steadily as the distance from Paris rises. But after the 500 km mark transportability plummets.
Another list using the AOC cheeses – that is those protected by a sort of patent – and a more precise location (at the département level; 89 départements in continental France) yielded an important information: no cheese is produced within a 50 km radius of Paris. Cattle was traditionally grazed in these regions and milk produced, as specialities such as the fontainebleau or the chantilly indicate, but it simply was not transformed into cheese.
The cheese belts
The specialties mentioned above are particularly delicate and have to be eaten more or less the very day of their production. This explains why they cannot be produced at more than a day-trip away from the consuming market (i.e. Paris). The same applies to milk, as von Thünen himself had remarked. As processing milk into cheese is a way to conserve proteins, there is no point to produce cheese any wherever milk does not require conserving. Hence there is a 50 km radius no-cheese ring around Paris.
A bit further – between 50 and 200 km from Paris – cheeses exist but are particularly soft. In fact, they are so difficult to carry that they usually come in straw baskets or in wooden boxes. This includes Normandy, the Parisian Basin and the Berry, which produce brie, camembert and numerous goat cheeses. Noticeably renowned cream and butter also come from these areas.
From 200 to 400 km, harder cheeses are produced; these are mostly the so-called washed crusts and stilton-like persillé types. They can be conserved longer and transported more easily, although many still require wooden boxes. These cheeses are for instance the epoisses, munster, dauphin and marolles from the Nord, Burgundy or the Alsace.
The most distant regions such as the Franche-Comté, the Alps, the Auvergne or even the Pyrenees are specialized in very long-lasting cheese such as the comté, the salers or the various tomes made of cow or sheep milk. A superfluous interpretation may explain the location of these long-lasting products by the mountainous profile of these regions, but a European overview of cheese production shows that it is rather a coincidence that in France most the hard cheese are produced in the highlands since many such cheeses in Europe come from plain areas such as the Netherlands, Emilia or Cheshire.
Finally, some areas manage to escape the Parisian influence. Most regions located further than 500 km (Provence, Languedoc) seem to have been crowded out by those closer from the city. Other closer regions are relatively unaffected by the Parisian demand possibly because of the lack of communication, this is particularly the case of Limousin. In other places, regional demand was certainly strong enough to compete with Paris. Lyon in particular managed to retain its soft cheese ring for its own consumption. I am particularly in the strength of this “cheese belts” theory as it closely resembles George Graham’s description of the patterns of grain production in early modern France.
What is the economic rationale behind all this?
The Dictionnaire universel théorique et pratique, du commerce et de la navigation informs us that, as late as the mid-19th century, only a fraction of the types of cheeses produced actually were commercialized at a national level, while others were only sold on a regional scale and the majority limited to domestic or at best very local consumption.
There seem indeed to be a double trade-off involved here. Although long-lasting products perform cheeses’ primary function (preserving proteins) much better, they appear to be much more difficult to fabric. For instance, the parmesan can literally last for years, but even experienced professional makers had to throw away part of their production because some mistake had been made somewhere. Brie on the other hand could not be preserved for more than a few weeks but every farmer could produce it with only variations in quality.
Nowadays, all cheeses have around 40 or 50% of fat content, but it seems that before the mid-19th century, most hard cheeses were significantly drier and their fat content was lower. Within the same type of cheese, brie for example, the fatter cheeses sold for significantly higher prices. So producers traded nutritional qualities for the possibility to preserve their products longer.
In areas close to cities cattle would compete for land, labour and other production factors with vegetables, fruits, poultry, etc. Further away, it would compete with firewood and most importantly cereals. So farmers in those regions had little incentive to specialize in cheese production. Cheese-making seems to have been more of a part-time by-employment than the core of their business. The production system was thus quite simple, no major dealer seem to have been involved in this trade; it may be that some producers even sold directly their cheeses on the markets.
Production was limited – symptomatically soft cheeses tend to be quite small – and very localized. In the vicinity of Paris alone the Dictionnaire universel listed no less than eight soft cheeses with roughly similar qualities (camembert, montherly, brie, bondon, neufchâtel, isigny, rougeret, macônnais). This high variability can easily be explained by the lack of competition (Paris could only import its soft cheese from a 200 km radius).
Hard cheeses on the other hand tend to be very large (up to 50 kg) and could be preserved for a long time, thus were easier to transport. In this respect their production could take place wherever competition from cereals and other products was not too strong, that meant generally away from the cities. The apparent exception of the cheeses from Holland is actually an excellent example: it is only because the Dutch countryside did not have to provide cereals to its large urban market (which relied on Baltic grain) that it could produce vast quantities of cheese. Stephan R. Epstein actually considered the development of hard cheese as part of the late medieval commercial revolution, as they greatly facilitated trade. Hard cheese typically traveled in boxes of 100 and even 200 kg, making them the equivalent of what barrel had been to the amphora.
These regions could thus specialize in cheese production. The Dictionnaire universel makes clear that cheese was very nearly the only export of rocquefort- or gruyere-producing regions. Specialization also allowed much greater production, in the 1850s the duchy of Lombardy could count every year on one million francs worth of fiscal entries thanks to cheese exports.
Economies of scale also implied new production systems. In the Milhau area, individual farmers would start processing the roqueforts but cave-masters would buy these unfinished cheeses to age them in special underground facilities. These few capitalists had a monopsomy on the sale of the very sought-after roquefort. They enforced their monopole on the farmers’ semi-finished cheeses with loans and by buying their products sometimes several years in advance. Unsurprisingly roquefort production was closely in tune with a series of spring fairs which allowed to export the production, sometimes as far as England.
In Switzerland, larger farmers dealt directly with merchants coming from the whole of Europe. Lesser producers gathered in cooperative cheese-making facilities called fruitières. Unlike highly-localized soft cheeses, hard ones were widely copied, their production spread from a region to another with the newcomers trying their very best to be identified and even mistaken with the traditional producing areas. That is why the Dictionnaire universel warns its readers against the cheap French copies of Swiss gruyere. Sometimes, the copy was well done and no one could make the difference, for instance in England, chester from Shropshire fooled many connoisseurs. Sometimes the new region was so successful that it competed for the very name of the cheese, for instance the Dictionnaire universel had to precise that many knew the parmesan under the name of “lodesan” (i.e. from Lodi). Finally, cheese recipes did not simply swept from a region to another neighbouring one; some producers copied successful yet distant cheese such as the Auvergne cheese-makers producing low-quality fake Dutch cheese.
So far, we have talked about cheese geography in terms of high transport costs, but evidently this is not the case anymore, so when did the traditional French cheeses location we can observe today mattered and finally settled? It is likely to predate the rise of railway transport, as the description from the Dictionnaire universel confirms. On the other hand, Abelor’s estimates of travel time in 1710 have a significantly lower explanatory value than sheer distance, suggesting that the origins of the French cheese geography is to be found somewhere between 1720 and 1840.
Arbellot’s “great alteration of the French roads”, which cut travel time by three or four between 1730 and 1750, is a likely candidate. Unfortunately the very strong explanatory value of the 1765 travel times was not statistically relevant and more research would be needed here using larger and more detailed samples.
Many cheeses still produced today pre-date this period, but what this estimate explains is not so much their appearance than their survival. It is also possible that some regions were highly specialized in the same cheeses as those they are producing today long before the mid-18th century. Philip H. Hoffman for instance indicates such specialization in 16th-century Normandy. But it is only normal for some regions to have been ahead of the curve.
It is also normal for the regions with a particularly strong comparative advantage (proximity from Paris, absence of competing crops) to have specialized early on. While the more distant ones may not have been able to count on regular communication with Paris before the end of the long string of catastrophes that plagued France’s economic development (the Hundred Years War, the Religious Wars, the long series of conflicts under Louis XIV). No doubt as well that the improved market integration of this critical product (aptly nicknamed in Lille the “meat of the poor”) did a lot to help 18th-century France to escape the daunting Malthusian (mouse) trap.
In France – and certainly in the rest of Europe – cheese production thus seem to be still influenced by pre-railways factors and in particular 18th-century travel time. Why subsequent improvements in transport technology did not affect this geography remains unclear. We can thus still observe “cheese belts” around Paris as cheeses become gradually harder, easier to carry and longer lasting as distance increases. The least commercially integrated southern and western regions, on the other hand, mostly produce soft cheeses originally designed for domestic and local consumption. Some towns such as Lyon managed to successfully compete with the Parisian demand and to preserve their neighbouring “soft cheese ring”.
We can thus say that unlike what its medieval and traditional trademarks often suggest, cheese is a product shaped by the market in a relatively recent period. So instead of being one of the symbols of the so-called altermondialist movement, cheese should be adopted as a totem by the market-friendly modernists.
ReferencesARBELLOT, G. (1973) “La grande mutation des routes de France au XVIIIe siècle”, Annales. Histoire, Sciences Sociales, 28/3, 765-791. “Fromage”, in Dictionnaire universel théorique et pratique, du commerce et de la navigation, t.1, 1859-61, Paris: Guillaumin, 1278-1285.
GRANTHAM, G. (1997). Espaces privilegies: productivity agricole et zones d’approvisionnement urbains dans l’Europe pre-industrielle. Annales: Histoire, Sciences Sociales 3, pp. 697-725. HOFFMAN, P. (1996). Growth in a Traditional Society. The French Countryside 1450-1815. Princeton: Princeton University Press.
KRUGMAN, P. (1991). Geography and trade. Leuven and Cambridge, MA: MIT Press.
MORICEAU, J.-M. (1994). Les Fermiers de Vile de France XV-XVIIIe siècle. Paris: Fayard.
MORICEAU, J. M. (2005), Histoire et géographie de l’élevage français : XVe-XVIIIe siècles, Paris : Fayard.
VAN ZANDEN, J. L. (1993). “The regional pattern of agricultural development, 1500-1800”, Paper presented to the Table-Ronde on L’évolution des rapports entre les regions de l’Europe, 14e-18e siecles, March 1993. Paris: Maison des Sciences de l’Homme.
VON THUNEN, J. (1966 ). Der Isiolerte Staat in Beziehung auf Landwirtschaft und National-okonomie. Darmstadt: Wissenschaftliche Buchsgesselschaft.
Benjamin Guilbert, this 11/03/09, Rabat