Its always fun in the course of research to find little pieces of information that are not related to what you are doing. One such example comes from my research in the Quebec Mercury to complete price indices past 1760. I found this little article concerning the coins minted in England from the reign of Elizabeth to 1805, organized by sovereign. For those who are interested, I have reproduced it here.
As I am nearing the completion of my doctoral thesis, I have begun to submit articles and share my data concerning the economy of French Canada prior to 1760. Obviously, this means that I should put up a “disclaimer” regarding the data as well as details regarding that data. Those can now be found here.
—PRICES AND WAGES IN NEW FRANCE, 1688 to 1760—
All prices are collected from account books of the Séminaire de Québec and from the Ursulines de Québec. Wages are also collected from these sources. All prices were reported in monnoye du pays – an accounting unit of money that was 25% inferior to the monnoye de France until 1720, after which date the two became identical. The prices of imported and manufactured goods were very often reported as the goods imported by the congregations for the benefits of their pupils, priests and nuns. The prices of agricultural and domestically produced goods tended to stem from the accounts related to the farms, estates and mills owned by these religious congregations. Overall, there were 27 goods for which reliable estimates of prices could be derived: wheat, flour, hay, bran, oats, eggs, socks, salt, codfish, peas, nails, shoes, savage shoes, lighting oil, firewood, wood planks, tobacco, beef, butter, lard, religious candles, wine, spirits, cloth, olive oil, soap, and eels. There were other goods, such as pepper, coal, wax, paper, molasses, glass panes, barley, sugar, guildive (rum from the French West Indies), vinegar, and iron, that were not reported continuously enough to be included in the construction of this price index, but that could, nonetheless, also be instructive. From 1688 to 1740, there were 1,431 possible data points for the above-mentioned 27 goods. Out of these data points, only 48 were missing, i.e. 3.35 % of the total. Most came from tobacco (8 missing), while the rest were evenly distributed. From 1740 to 1760—years marked by war and inflation as a result of monetary policy—there were 20 goods that were reliably collected and that represent a potential for 420 observations. Out of these, 70 data points were missing and needed to be imputed, i.e. 16.7% of the total in this 21-year period. The method used to compensate for missing data points was as follows. I took the price of these goods relative to that of wheat per five-year period (except for the periods 1688 to 1694 and 1755 to 1760) and took the average over the period to complete the gaps. The interpolation for candles and cloth was different, however. The majority of the missing observations past 1740 come from these two goods. Candles prices were accurately reported from 1688 to the 1720s. After that point, the observations become fewer and farther between, but observations for tallow continue. Simultaneously, we have long series of prices for religious candles (cierges). The difference between the chandelle and the cierge is that the former was made out of tallow, whereas the second was made with yellow wax, which was more expensive. Throughout the period, their evolution overlaps and they move quite similarly (at different levels). In the years where data was obtained for both goods, the price of candles averaged 19.92% of the price of cierges. Since tallow, candles and cierges are all products that can act as the basket of illuminates, we relied on tallow and candles (although the movements of cierges are similar, which is reassuring) to create a unified series. With regards to clothing, prices are reported sporadically after 1740. This is not the case with socks and shoes, which continue to be reported with reliable frequency. From 1688 to 1740, the price of clothing moves in the same broad direction and amplitude as the price of shoes and socks. Hence, the movements of sock prices past 1740 are given to the price of cloth.
Prices for the same months of every year were not available, so I recorded observations for all goods and services found, and a maximum of five price quotations per year were recorded. The reason for this is that, in numerous instances, prices were reported but not the month of the transaction. Although imperfect, this method of collection compares very well with the few existing datasets of prices. For example, I continued collection of wheat prices up to 1770 in order to compare with the source provided by Ouellet, Hamelin and Chabot (1982), which starts in 1760. The prices line up nearly perfectly and move in the same direction, with the exception of the year 1760. The same happens in terms of the movement of raw cloth with the prices reported by Egnal (1998) and with the prices for firewood collected from Richard (1973). Although the vast majority of the data comes from the more detailed and richer sources associated with the Séminaire de Québec, it is worth pointing out that the prices for the two sources move in the same way and at the same level, which points to the fact that using the two sources in combination is not a problematic approach.
Measurements were a considerable problem. Rousseau (1983) provided equivalences in his work on the Augustine congregation that ran the hospital in Quebec City. These are illustrated in Table 1 below. One should be very careful not to confuse the French livre and the English pound. Although the two are linguistically identical (livre in French means pound in English), they represent different weights in grams. The French livre (translation for pound) carries 489.5 grams, against 453.6 grams for the English pound, a 7.91% difference.
Table 5: Bare bones welfare ratios between Nova Scotia and New France
|TYPE OF ITEM||UNIT USED FOR THIS PAPER||OTHER UNITS AND THEIR CONVERSION RATES|
|Liquids (spirits, wine, burning oil)||Pot (one pot = 1.9 liters)||2 Pintes = 1 pot
Barrique = 120 pots
Baril = 35 to 40 pots
Velte = 4 pots
Quart = 50 pots
Tonneau = 4 barriques
|Liquids (olive oil)||Livre (489.5 grams)||Baril = 80 pounds
Pot = 1.9 liters
1 liter = 967.88 g
|Grains and others (wheat, oats, bran, flour, peas, salt)||Minot (one minot = 1.1107 bushels)||Setier = 156 liters
1 minot = 39.25 liters
1 minot of wheat = 60 livres
Pochetée ≈ 1 minot
1 barrique of salt = 6 minots
|Lard and beef||Livre (489.5 grams)||Baril (lard) = 150 livres
Baril (beef) = 200 livres
|Codfish||Poignée (one poignée = 2 codfish)||Quintal = 72 codfishes
Average net weight of a codfish = 5.107 livres
|Cloth||Aune||Aune = 1.3 yard
Yard = 0.9144 meter
Source: François Rousseau. 1983. L’œuvre de Chère en Nouvelle-France : Le régime des malades à l’hôtel-Dieu de Québec. Québec Presses de l’Université Laval, p.394-396; Christoph F. Grieb and James Oehlschläger. 1857. A Dictionary of the English and German Languages, with a Synopsis (…), Vol 2. Philadelphia, PA: John Meif, p. 1132.
As for the wage data, figure 1 illustrates the distribution of the data collected. Monthly and annual wages are not discussed here because they are few in numbers and the nature of the work for which they were given is not always specified. Moreover, they were often indentured workers with payment-in-kinds that were not specified. The daily wages are much cleaner. Very rarely were those wages associated with payment in kind, so we have very little problems arising from underestimated wage levels. When they were, the account books added the notice of “et nourry” (and fed) to the wage rate, or mentioned a specific item that was offered. Some of the tasks of the workers were identified through their surnames (Masson meant mason, while Meunier meant miller). Most of the time, the tasks were specified. When they were not, it was possible to use the Séminaire’s indexed name cards to find who these individuals were, allowing us to complete the information about each observation. Like the prices, the wages were reported in monnoye du pays. The early data for carpenters is missing for 1688 and 1689. The value observed for 1690 was attributed to these. The same problem was observed for 1688 with unskilled wages. I attributed the 1689 value. All other subsequent interpolations, which are very rare for unskilled workers but not for carpenters, were made using the MATLAB software and the “iterpl” function from the set of the 1-D interpolation methods. The option “spline” was used for the method of interpolation. This option represents the piecewise cubic spline interpolation method as in de Boor using not-a-knot end conditions. The usage of the piecewise treatment helps to avoid Runge’s phenomenon of oscillation of polynomials of higher degrees and decreases the interpolation error compared to linear or simple polynomial interpolation methods (de Boor 2001:23). The interpolation has been kindly provided thanks to the help of Vadim Kufenko of Hohenheim University’s department of economics.
Prices for England, Paris and Strasbourg were collected from Allen (2001). The same is true for wages from England and Paris. The same original source for Strasbourg was used (Hanauer 1878), but with the substantial modifications discussed above. The wages for Boston were taken from Main (1994) and converted into grams of silver using the silver conversions proposed by Lindert and Williamsom (2014), which yielded similar rates of wages as the silver conversion for London proposed by Allen (2001). A greater discussion of the Boston wages is provided in the text. The wage rates for Philadelphia are derived from Gary Nash (1979: 392-394) and the values for 1720 to 1726 have been interpolated using the 1727 value. For Boston and Philadelphia, we used the same methods and sources of interpolations and computations as Allen et al. (2012), barring the following exceptions: the price of maize was computed the same way that Allen et al. did. The price of wheat was interpolated according to moving average of two periods were missing up to 1721. After 1721, wheat prices for Boston follow the wholesale price movements proposed by Arthur Cole (1938). We used the same weight conversions as Allen et al. (2012). For firewood, there would be too many missing observations had we used Allen’s method. We proceeded by taking Cole’s data for staves in Philadelphia (available after 1721) assuming that prices for firewood follow those of staves after 1721. Prior to 1721, linear interpolation was used. That series was used both for Boston and Philadelphia. Allen and al. did not possess clothing prices prior to 1747 in the American colonies. They assume a constant price up to 1753. They mention that “most cloth was imported from England and could be transported to Massachussets, Pennsylvania and Maryland at similar cost” (2012: 21). Yet, the price level they compute seems to be somewhat too high. Colonists tended to wear mixed garments of clothing, whose prices tended to be lower than those suggested. The historian William B. Weedon (1890: 890) suggested that in 1713, one yard of plain cloth and one yard of checkered shirting all sold at 1.25 shillings. Given that one yard represents 0.9144 meter, it means that the price per meter stood at 1.367 shillings. Using the exchange rate of 3.7127 grams of silver per shilling proposed by Lindert and Dietrich, this means a price of 5.08 grams per meter. This price for 1713 was indexed to the price movements of linen in England using Clark (2005). This results in a price for clothing closer to the reality of consumers.
Finally, with regards to the conception of the baskets, one would normally have preferred to include bread, but bread prices were not available for New France. Hence, we fell back on wheat and adjusted the basket as a result. The conversion of calories from minot of wheat to kilograms followed table 1 above and the energetic values supplied by Dessureault (2005: 265). A minot of oats was 34 pounds, while a minot of wheat weighed 60 pounds. But a minot of wheat yielded 96,728 calories and one of oats yielded 60,509 calories. Considering that this is a unit of volume of 39.25 liters, these did not necessarily have similar weights. One minot of oats weighed 34 pounds, one minot of wheat weighed 60.7 pounds and one minot of peas weighed 60 pounds. This meant 395.5 grams of oats per liter, 706.16 grams per liter for wheat and 698.01 grams per liter of peas. An adjustment for losses in the transformation of wheat into bread has been factored in, which explains the higher wheat figure in the respectable basket. The bare bones assumes that oats was simply eaten as porridge (known as gruau in New France), which is why the quantity is not adjusted.
Allen, Robert. 2001. “The great divergence in European wages and prices from the Middle Ages to the First World War” Explorations in economic history, Vol.38, no.4, pp.411-447.
Allen, Robert, Tommy Murphy and Eric Schneider. 2012. “The Colonial Origins of the Divergence in the Americas: A Labor Market Approach”, Journal of Economic History, Vol. 72, no. 4, pp.863-94.
de Boor, Carl. 2001. A Practical Guide to Splines (Revised Edition), New York: Springer.
Clark, Gregory. 2005. “The condition of the working class in England, 1209-2004”, Journal of Political Economy, vol.113, no.6, pp.1307-1340.
Cole, Arthur H. 1938. Wholesale Commodity Prices in the United States, 1700 -1861, Statistical Supplement, Actual Wholesale Prices of Various Commodities. Cambridge, MA: Harvard University Press.
Egnal, Marc. 1998. New World Economies: The Growth of the Thirteen Colonies and Early Canada. Oxford: Oxford University Press.
Hanauer, Charles Auguste. 1878. Études économiques sur l’Alsace ancienne et moderne (denrées et salaires). Paris : Société industrielle de Mulhouse.
Lindert, Peter and Jeffrey Williamson. 2014. American Colonial Incomes, 1650-1774. Cambridge, MA : National Bureau of Economic Research.
Nash, Gary B. 1979. Urban Crucible: Social Change, Political Consciousness, and the Origins of the American Revolution. Cambridge, MA: Harvard University Press.
Ouellet, Fernand, Jean Hamelin and Richard Chabot. 1982. “Les prix agricoles dans les villes et les campagnes du Québec d’avant 1850: aperçus quantitatifs” Histoire Sociale / Social History, vol. 15, no. 29, pp. 83-128.
Richard, Claude. 1973. L’industrie du bois en Nouvelle-France au 18ème siècle. Master’s Thesis, Department of History, University of Montreal.
Rousseau, François. 1983. L’œuvre de Chère en Nouvelle-France : Le régime des malades à l’hôtel-Dieu de Québec. Québec Presses de l’Université Laval.
Weedon, William Babcock. 1890 . Economic and Social History of New England, 1620-1789. London: British Library Historical Print Editions.
I have a new working paper co-written with Germain Belzile – it is by no means complete. I intend to make some corrections to improve it further. It concerns an important chapter of Quebec history: the era before the nationalization of electricity. We contest the claim that the companies were acting as monopolies and gouging consumers. We find that electricity prices were dropping, output was increasing, productivity was increasing and that was in spit of a greater demand caused for Quebec electricity exports by Ontario which subsidized consumption. We don’t pretend that everything was “honky-dory”, we argue that the case for nationalization was very weak relative to private production. Here is the abstract and the paper is available here.
Abstract: Upon opening history books about the electrical industry in the Canadian province of Quebec prior to nationalization (which was realized in two steps between 1944 and 1962), one is often confronted with the claim that the industry was monopolistic and was gouging consumers via predatory pricing – especially when compared to the neighbouring province of Ontario. All though it is hard to collect price data at the level of firms, it is possible to collect some overall – but often ignored – data about the industry to evaluate this claim. With the use of such data over time, we can observe the opposite. Rather than behaving as oligopolies, the electrical firms in Quebec increased production faster than elsewhere while prices did the same in the period before nationalization. Moreover, there is strong evidence that productivity growth was higher in Quebec than in Ontario and the Canadian average. The main reason for this divergence between facts and history books is most likely the choice of benchmarking Quebec with Ontario – a province which had opted for public production in the early 20th century.
In a recent paper, submitted to Enterprise and Society, Mathieu Bédard and I argued that free banking in Canada boosted economic growth. However, we proposed another channel for this effect than the proper inter-mediation of savings and investments. We argued that private money produced by banks who backed their notes with their own assets allowed “good money” to crowd out “bad money”. Basically, the “bad money” was the low (and very uncertain) quality of hard currency. Before the appearance of those banks, “good money” circulated at a premium. We argue that private banks solved this important and simplified trade within Canada.
While doing research for other papers, I stumbled upon an article in Le Télégraphe (a Quebec city newspaper) which actually discussed the problem of poor quality currency and how private currency was actually solving the problem. The image of the article (in French from the June 3rd 1837 article) is seen below. I am always sad (and happy) to discover such interesting pieces of supporting evidence after I made my case.
Oh well, at least I know me and Mathieu weren’t wrong!
I have been experimenting with new data for a paper I am writing (alongside Vadim Kufenko and Alex Arsenault Morin) regarding mortality and land tenure laws in Canada prior to confederation. Interestingly, I had on my hands a new dataset (which I had collected over 6 months ago) of prices and wages in Lower Canada (Quebec) for 1831 which offers a broad cross-sectional of grain-wages (what quantity of wheat one could buy with one day’s work). By accident, I stumbled across documents offering deaths and births for the same cross-sectional decomposition. I have been playing around with the data and I have found some very interesting relations.
The most interesting is that, although unsurprising for quantitative historians, mortality was negatively related with wages. The higher the wage rate, the lower the mortality statistic. However, this obvious fact is made fascinating by another fact: grain wages are strongly related with institutions. In Lower Canada, two legal regimes co-existed regarding land tenure. The first was based on French law and was contained to lands settled prior to 1791. The second was based on British law and all new settled lands would be settled under this law. The system based on French law was feudal in nature and landlords had monopoly rights and duties they could charge on peasants who would never own the land plots granted to them. The system based on British was pretty much “homesteading” in the sense that once you had staked out a claim, it was yours (if you managed to get it registered). For the fun of working of data (and before I try to flesh out a more substantive article – I have to finish the ones I am doing with Kufenko and Morin and some others), I just regressed the log of grain wages to the land tenure system with some controls I created for another paper (distance from urban areas, land quality). The result is that we end up with higher wages in areas under laws by 22% and the gap is significant at the 99% level (and the R2 is 14%).
Obviously, more work should be done but if wages were depressed by French law and wages were related to mortality, then French seigneurial law was related to the mortality rate. More to come in a few months!
One of my papers, co-authored with Vadim Kufenko of the University of Hohenheim, has received the final green light for publication in the Journal of Population Research. The paper is titled : Malthusian pressures – empirical evidence from a frontier economy.
This is, I believe, a very significant paper for Canadian historians. Simply put, it uses new price data to measure the effects of changes in real prices upon mortality and births to test for the existence of population pressures on economic performance prior to 1860 in Canada. Up until now, the dominant view has been that the largest colonies suffered from “too much people for too little fertile lands”. In short, the dominant view was “Malthusian” whereby more population means lower real wages and more poverty. The Malthusian prediction is that fertility should drop and mortality should rise as a result of too much population (put very simply here, but expounded in more nuances in the paper). Kufenko and myself contest this narrative by showing that the traditional stipulations and predictions of population pressures did not materialize. We propose that improvements in productivity and economic organization had largely beneficial effects that countered the Malthusian pressures. The result throws into doubt the claim that Canada suffered a deep and prolonged economic crisis during the first five decades of the 19th century and it eliminates the Malthusian channel as an explanation of economic events in early Canadian history.
The paper can be read here at ACADEMIA until it is published formally in the Journal of Population Research
***Note: I wish to thank the anonymous reviewers, my co-authors and the participants of seminars at HEC Montreal and the University of Hohenheim for their comments.
Alongside Vadim Kufenko (partner in all crimes) and Alex Arsenault Morin, we have organized the prison records of Quebec City between 1813 and 1848 to measure the heights of the prisonners. Our aim is to study the presence of differences in biological living standards across ethnic lines over time in Quebec and use this measure as an crude indicator of inequality and overall living standards. The paper is at present a short note (a few thousand words) which we have submitted to different conferences (Economic History Society and Economic and Business History Society). We will lengthen the note to a paper once comments have been provided.
In the meantime, here are the preliminary results illustrated and the abstract of our submission to EHS and EBHS conferences is here.
The figure shows the heights in inches by year of imprisonment of males above 20 years of age according to ethnic origin.