May 22nd internal meeting

Present: LoÏc Charles, Guillaume Daudin, Béatrice Dedinger, Paul Girard, Demba Konate, Laura Moutou

1) RICARDO : presentation of the work by Paul, Béatrice and Matteo Azzi
It uses opensource code (deposited at github). Licence OpenSource (interdit d’inclure le code dans une application dont le code est fermé)

3 tools
– Database creation : from Access to SQL Lite (MDBTools did help — a number of data correction were made as well)
But Access has been kept as the reference database.
– Client : web interface in HTML 5, using Java D3JS by Matteo Azzi
This interface can be scripted, but one has to understand the API language (and OpenRefine should help)
– API : links between the client and the database

We have left aside for the time being the time-dependant colonial dependencies : we will need more funding for that.

We hope this work will be useable for TOFLIT18 afterward — still we will need one month of designer’s work.

2)  We discussed the TOFLIT planning and how to organize Guillaume Plique’s contract. We will move him from eight to seven months, and have him start latter, till early 2016.

3) Online storage of the data
We will try LiveDrive
Nakala from Humanum is not adapted at all to what we are trying to do.

4) Call for tender for transcription
Everything is done. SMSI (national data) should be ready in mid-july. TASECRA (regional data) in two batches : one in mid-september, one in mid-december

5) Laura and Demba introduce themselves

6) We have to obtain « Etats de dépenses » from our administrations

7) Next meeting : July 3rd at 10. All are welcome !

Report on the May 23rd workshop


Thank you to Alexandre Aubourg, who wrote this report on the May 23rd meeting on Early Trade Statistics at the OFCE, Paris

For more details on the working day, see

Attendance :

The first presentation was a welcome word by Jérôme Creel, who remined us of the economic history tradition in the OFCE.

Loïc Charles (University of Paris 8/INED) gave a presentation on the findings and the issues raised by the questionnaires on trade statistics. The statistics accumulated can be classified according the geopolitical unit to which they belonged: ports, regions or nations. While each unit has its advantages and disavantages, another geographical issue brought to light was the presence of a land frontier.The collection of the data was carried out for fiscal purpose even though mercantilism was on the wane (the rise of the Weberian state comes as a possible explanation). Some technical issues come from prices – which are sometimes fixed or lacking. The creation of common measures between monetary accounts, weights and quantities or the use of premiums added to a given set of price may help solve these problems. While bilateral comparisons could be interesting, one should be aware that any economic issue will be constrained by the dataset available. It was also underlined that when indications of destination/origin, price, long series are present, it would be quite interesting to compute indicators of trade specialisation. Some programmatic issues introduced : one may adopt a comprehensive approach involving the names of individual goods and a description and translation work (which was prefered by the audience) or a limited approach using the value of trade per category expressed in common currency.

Ulrich Pfister (University of Münster) presented a paper on import trade in Hamburg (1733-1798), the port of which had key role in the German trade overseas- notably the colonial trade. The dataset is composed of 187,000 declarations of value of imported goods coming from import toll ledgers and provides a consistent picture of Germany’s international economy. For simplicity, 40 major groups were notably obtained from 6,900 terms. Moreover, deflating current Hambourg price allowed to obtain prices for 44 commodities of interest. The results show an increase in openess, while no steady trend and war-related schocks can be observed. However, the dataset is flawed by under registration that may be explained by transit trade and rival satellite ports.

Werner Scheltjens (University of Leipzig) presented a study of French import for the Baltic between 1670 and 1850. The main source are the Sound Toll Registers, which are the records of the toll levied on the passage of ships through the Sound. A significant part of the work consisted in homogenizing and the standardizing products, converting weights and measures and combining them with origins. The results notably highlight a sharp decline in the French share of imported goods in the late 1770s.

Cristina Moreira (University of Minho), Jari Eloranta (Appalachian State University and University of Jyväskylä), Jari Ojala (University of Jyväskylä) and Lauri Karvonen (Åbo Akademi University) discussed over methodological issues in comparing trade flows between Portugal and Scandinavia, which proved to be very good partners. One major issue regards the use of values (which may be easily compared but leads to difficult problems to deal with) and volume (which are precise but not easily comparable). Ulrich Pfister argued that one should work with real volumes and create an index of trade to get the evolution of prices. As it was also pointed out during the discussion, the use of bilateral trade, which does not always represent the genuine structure of trade, may also be questionned. We also face matching problems, which may lead us to focus on ships.

Jeroen van der Vliet (Rijksmuseum Amsterdam), whose aim is to link Amsterdam and Sound data sources, showed how he used a shipment approach and looked at the size of Dutch merchants fleet and the tonnage per destination. His main source are the “musterolls”, which provides a comprehensive biography of the ships. After describing the usages of the merchants in Amsterdam and comparing the data, he discussed the link between the utilization of the flag of convenience and the apparent relative decline of Amsterdam over the 18th century.

Hubert Escaith (WTO) gave a presentation on past and current issues in trade statistics. After a short talk about the evolution of official statistis from the 18th century, the frequentist and subjectivist schools and their success over space, sectors and time were presented. Hubert then launched in a discussion on current issues in trade statistics – notably the motives for collecting trade statistics (which are useful for the government – among other things to show their solvability, as François Mouriaux pointed out) and the importance of tracking value-added trade.

Guillaume Daudin (Univerisity of Paris-Dauphine/OFCE) presented a comparison between the French Balance du Commerce and the Sound Toll database. First, the sources have to be made comparable, which means that some issues related to the classification of goods, translation to French, quantity units but also geographical changes have to be fixed. Comparisons of the sectorial composition of trade in a given year, the volume of trade, the values, the number of pieces and the longitudinal evolution of trade throughout the period are then carried out. While some anomalies regarding the composition of trade were highlighted, it was found that the composition of Sound taxes was related to the composition of the value of French trade.

May 24th meeting

Thanks to Alexandre Aubourg who prepared the first version of this report on the May 24th meeting.


Present : Alexandre Aubourg, Bertrand Blancheton, Loïc Charles, Guillaume Daudin, Béatrice Dedinger, Pierre Gervais, Demba Konate, Ivan Ledezma, Yannick Le Pen, Silvia Marzagalli, Laura Moutou, Pierrick Pourchasse.

1) Some issues related to the German and Italian sources :

  • The contrast between German and Italian sources is striking. We have found many specialists for Italian sources. We could ask Biagio Salvemini ? Maria Furaso ? If we need more.
  • Regarding the German sources, we do not know who should be contacted. Maybe is it the case that little German sources survive ? There is no secondary sources before 1830s and the existence of primary sources is called into question. Werner has worked on Stettin and many documents are not available. Loïc will ask Christiane Lebeau. Regarding the Prussian case, Pierrick underlined that sources for merchandise per port coming from the Danish consuls should be available.

2) OFCE review and questionnaires :

  • We are thinking about a general description of the sources, a lexicon and a presentation of the data (eg the flag issue).
  • During the OFCE meeting, we noticed a discrepancy between quantities and values (but the questions came mainly from economists).
  • The questionnaires should be standardized and some points have to be made explicit.
  • A few pages on the difficulty to handle different sources could be written.

3) Transcription of the data

  • The national data will be retranscribed by SMSI by the end of the summer, while the transcription of the regional data will be carried out by Thasectra (a large portion should be ready for the end of the year).
  • Colonial data are mainly located in Aix-en-Provence (used by Tarrade) (a small portion may be found in Paris in the Lavoisier archives, but it is probably not very interesting). We are notably interested in obtaining data that is disaggregated by colony.
  • Discussion on how they computed the « national » price

4) Classification of the data

  • The possibility to incorporate Silvia’s and Pierre’s work (classification in 3 dimensions) was highlighted.
  • We notably have to fix the issues related to the meta-categories and the internal categories.
  • We also discussed the ties with Blancheton’s work on the 18th century and the goals of Huma-num.

4b) Other link with Navigocorpus

  • As soon as 1787 is ready, we will compare it with Navigocorpus’s results

5) Ricardo 1830-1938

  • The project was presented by Beatrice. The main objective is to put the database online and the first version is planned for the last quarter of 2014. The data base is being cleaned up for one year and some issues regarding the country classification were raised.
  • We discussed the wisdom of integrating IT people in our projects, and contrasted Navigocorpus’s choice with TOFLIT’s and RICARDO’s

6) Calendar

  • The importance of the intern workshop planned in november/december 2014 was underlined. A reflexion on the tools, the strategy for interpretation and maybe a special issue should be launched.
  • We also addressed the possibility to set up a European project in the near future.

7) An application for Huma-Num accreditation will be submitted in september

  • We also mentioned the possibility to create a consortium.We may first look for technical solutions in order to enter the data. Then, these ressources will be made available and participatory research will be encouraged. The valuation and the reflexion on the comparison can be done beforehand.
  • Bertrand will see Bourdalé in early June
  • Other projects : Fichoz //
  • We discuss the difficulties of these projects and their long-term viability