February 3rd 2017 meeting

Attending : Loïc Charles, Guillaume Daudin, Matthias Loise, Florence Perret, Guillaume Plique

1. Paul Girard & Sciences Po
Paul is discussing his status with SciencesPo.

2. Guillaume Plique’s planning
We have six confirmed weeks of work :
February 6th and 13th (2)
April 10th, 17th and 24th (3)
May 1st (1)
We have funded 16 weeks of work. We need to plan for the other 10 weeks ? 4 weeks in May, 4 weeks in June and 2 in July seems a bit extreme.

3. Simplification of goods
The works has been completed. Pierre Gervais is going to proofread Florence’s work on the datascape.
Pierre suggested to go further into the simplification, gathering all kinds of ‘agaric’ for instance. The group decides to stick with the original idea and not loose any information at the simplification stage.
During further discussion, it is decided that Pierre will not spend three weeks making sure that all simplification rules have been perfectly applied.

4. Looking for mistakes in the data
Matthias Loise has made a linear regression on the prices using product names, metric measurements unite and year. This way, he could forecast the prices and treat those which are too different from their predicted values. The rate of mistakes is still high. With the residual between 2,5 et 3, there is 60% of « true wrong prices », 20% of « actually right prices » and 20% of doubtful cases. The share of true wrong prices is higher in local data (70%) than in the national data (55%).
As for now he could only work on metric measurements but it could be interesting to work on other units although it’s going to be hard to make comparisons (as those units are not standardized especially those emanating from the colonies).

5. Data: Assistance Publique & Marseille
Loïc Charles will look for new data mentioned to us by François Velde at the Assistance Publique (fond Montyon) during the following week. For now, he has been working on Marseille’s sources.
He basically found many small things, but no missing year.
– Imports per product in Levant’s échelles from 1725 to the mid 1770s (and some exports)
– Some colonial import and export trade statistics giving the destinations.
– Some documents coming from the beginning of the second bureau
– A list of freight costs at the bilateral port level

4. European funding
Loïc Charles met with Lucyna Derkacz/Gomez-Echeverri in charge of European projects at Paris 8. She was enthusiastic about Toflit18 and summed up the criteria to apply for a European funding. It appears that two programs could be interesting for Toflit:
– SSHS program. This depends of the work program and whether we fit in what they are interested.
– FET future & FET exchange programs (http://www.horizon2020.gouv.fr/cid110740/appel-2017-fet-open-coordination-and-support-actions-csa.html)
Those are relatively low budget programs which were created to finance interdisciplinary projects including new technologies and private companies, criteria which matches with Toflit. New adjudications are going to be made for 2017-2018.
Loïc Charles thinks that, first, we could duplicate the way we created the database to obtain similar practices, and then build a translation’s procedure and a multilateral database. In this way, we would be able to map not only one country but many simultaneous European perspectives. We could also match our data with maritime data (Sound Toll Registers http://www.soundtoll.nl/index.php/en/over-het-project/str-online & Navigocorpus http://navigocorpus.org/) but it would be a lot of work.
Guillaume’s contacts in Dauphine had been more interested in the Infrastructure program.
Guillaume and Loïc decide to each write a 1-paragraph defense of each option and send it around to the list to see who is interested in what.
Another option would be to create an ANR with European partners.

5. We discuss the scientific committee of the 2017 conference

January 5th 2017 meeting

Thank you to Florence for the first version of this report!

 

Attending : Loïc Charles, Guillaume Daudin, Béatrice Dedinger, Pierre Gervais, Paul Girard, Matthias Loise, Florence Perret, Guillaume Plique
1. Planning Guillaume Plique
He is one month late of the preceeding project. And starting in February he is starting on Wasmer’s project (with a workshop at the end of March).
In total we have nearly 4 month (10 700 € from SciencesPo and 13 000 € from INED) Planning:
– First 2 weeks of February
– One or two weeks in March
And the rest ?

2. Paul Girard & SciencesPo

3. Looking for mistakes in the data
Matthias Loïs has been looking for mistakes in the data
a. By looking at trade flows over 25 million l.t. A number of typos.
b. By comparing retranscribed trade flows and predicted (by a repeated sales model) trade flows. Only national trade for the time being. A lot of too-small trade flows that are not very interesting to correct.
c. Will be looking at the difference between retranscribed prices and predicted prices (by a repeated sales model). Many mistakes in Marseille
It is often because they forget to divide by 100 or they mix « sous » and « livres ».

4. Data
Marseilles’s data are a pain to check and integrate in the database. Loïc is doing it on his own and is spending a lot of time on each document.

This is blocking the integration of Dardel’s Rouen.

We decide to put Moyons at the Assistance Publique on the top of the pile.

5. Simplification of goods and measure units
Matthias has treated the measure units with three successive files.
The work on Merchandise Simplification has started again thanks to Florence. Pierre is going to participate and aims at finishing at the end of February or March.

6. 2017 Conference
Pierre makes two proposals.
– Use Chaurand and Gradis and check if their data is compatible with ours
– Study the link between the complexity of classifications and specialization between ports. Start with wine ?

7. Next meetings
03/02-9:30 : medialab
08/03-9:30 : medialab
21/04-9:30 : medialab
01/06-9:30 : medialab