Skip to main content
Funded by the ERC

EU Funded Project

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme under grant agreement No 724544

Average-Transaction Costs and Risk Management during the First Globalization (Sixteenth-Eighteenth Centuries)



Average Reports

Intellectually and scientifically, the documentation produced within ‘ Averages’ procedures confronts us with data which is splendidly encompassing in terms of the description of maritime trade, and which can allow its analysis at a level of granularity which is simply unprecedented.

General Average is structurally and substantially different from Insurance, not only it is older, but it has remained a strictly mutualistic form of protection and, unlike insurance, has not developed into a speculative instrument. I am therefore arguing that, notwithstanding strong commonalities in the cultural elements underpinning the evaluation of risk, and the way these directly affected the operational measures put into place for its management, GA – as a non-market phenomenon – is a more stable unit of economic analysis than insurance, as the latter was (and is) volatile, especially given its evolution into a speculative instrument. The quantitative data produced during GA procedures provides a different type of data, with values closer to the ‘real’ ones – that is to say, to the values perceived to be ‘real’ by all involved – simply because all who were involved in GAs were active participants in the business venture, and therefore over/underestimation of costs would have affected all parties, each with substantially different economic interests within the venture.

The reports produced by the interested parties when starting a GA procedure always provided data on: ship’s name and typology, master’s name, route (origin and destination), cargo composition, length of journey, narrative of the accident, and the list (and monetary value) of extraordinary costs or damages sustained. Not all original reports were followed up with a proper GA procedure, as in some cases the report was filed as a cautionary measure. For those cases when the initial report resulted in a proper declaration of GA, further quantitative data becomes available: value of the ship and cargo; damages/costs and sometimes details on freight contract. On the basis of this data, expenses were apportioned to the shipowners/merchants and then validated by the relevant local authority before payments were made.

As GAs originated in accidents during navigation, and these were primarily caused by bad weather, the ‘reports’ which are an essential element of GA procedure, provide detailed descriptions of such occurrences, and this will allow provision of data on the distribution of storms and extreme weather events. This data will be made available to the ‘Atmospheric Circulation Reconstruction over the Earth’ (ACRE) team members based at the MET office (in Exeter). Rob Allan has been involved since the beginning, to make sure the AveTransRisk data will usable by those investigating pre-19th century historical weather patterns.


Since the conception of AveTransRisk, I have been in constant dialogue with the Navigo (formerly Navigocorpus team whose leader – Silvia Marzagalli – has been brainstorming these issues with us since the beginning. The final aim is to have a dataset which can – potentially – be integrated with Navigo to increase its usability and time-resilience.

The latter is a serious long-standing concern of mine. Indeed the issue of ‘digital archives’ in itself will be a challenge for historians in the future. I have raised the issue at high level European research bodies, and although there seems to be some growing awareness of this, there is no clear common policy to invest in the technical infrastructure of ‘backward compatibility’ that is sorely needed.

Many Digital Humanities projects die with the end of funding, and very few institutions are taking this issue seriously. France and the Netherlands appear – at least for now – as a partial exception to this rule, particularly the CNRS and the Dutch CLARIAH are actively working to resolve some of these important issues.

The roots of the Database are in the work of the late Giuseppe Felloni, who since the 1990s collected data on general average procedures in Genoa, as he was interested in them as a tool to evaluate the economic performance and resilience of Genoese maritime trade and economy. As his research interests moved onto the Banco di San Giorgio and its crucial role in the history of finance, he moved away from GA and, at his retirement, bequeathed his GA material (10 boxes, about 2400-2500 datasheets) to the Library of the Centro di studi e documentazione di Storia economica "Archivio Doria" del Dipartimento di Economia AveTransRisk team member Luisa Piccinno is coordinating access to this material and its upload into the Database.

My major concern is to make sure that the data is accurately and cleanly collected and uploaded into the Database in the cleanest possible way. All the data populating the Database comes from the original documentation, and the data in the paper cards is being checked against the original before being uploaded. We have discovered that in the file cards there were some ‘estimates’ – e.g. about ship tonnage – and these will not be in the Database. What we shall provide, at a later stage, is the Silver equivalence, so as to facilitate future comparative analysis of the value and prices given in Average procedures.


I cannot praise high enough the splendid collaborative exercise which is at the root of the Database development, much of which goes on behind the scene through email exchanges, conversations and meetings – real and virtual. This is the best teamwork, and great evidence of how effective international collaboration can be.

The project is also benefitting from two external collaborations, which are greatly enhancing the project activities and results.

Malta: The Notarial Archives in Valletta are currently undergoing a programme of rehabilitation supported by a European Regional Development Fund grant. Thanks to an agreement with them, and their generous provisions of free access and help of volunteers, we have been able to digitise the large collection of Sea Protests and Average papers which were in the archives of the notaries William Stevens Sr., William John Stevens Jr. and Charles Curry. This is an exceptionally rich and coherent set of documents whose study is shedding light on the nineteenth century transition in GA practice. This data is currently being uploaded into the AveTransRisk database, thanks to the work of Maltese volunteers.

Livorno/Pisa: team member Andrea Addobbati, successfully obtained from the Italian Ministry for University and Research (Ministero dell'Istruzione dell'Università e della Ricerca) funds to activate a one year research assistant position on “Administering Justice in Trade: Mercantilisms, courts and merchants of Mediterranean Europe (XVIIXIX Centuries)”, based at the Department of Civilizations and Forms of Knowledge (Scientific Sector M-STO/02 Modern history). The winner of the national competition – Dr. Annalisa Biagianti – worked during 2018 on the data collection and analysis of those average adjustments stored in Livorno state archive. Additional material has also being collected by local volunteers in that archive, under the supervision of AveTransRisk team members.At the moment, due to the Covid-19 emergency, this work is suspended, but we hope to start again when the archive will reopen to the public.

Documents and Data

The material is both consistent in its structure, and varied in its morphology. There is a great operational convergence, as declarations prepared in one country needed to be valid in others, and we shall soon upload some texts which shall describe the peculiarities of each of the locations where we are collecting data.

However, the documents at the basis of the data belong to two main types:

  1. Normal procedures  of ‘averages’

  2. Litigation cases

    Both types are easily distinguishable in the Database, as the origin of the documentation needs to be taken into account before interpreting and analysing the relevant figures.

We are also developing interactive maps, so as to give visual representation of maritime trade routes, and provide the location of extreme weather events.

(Maria Fusaro)

Technical Challenges

The AveTransRisk project required a means of capturing the source material from remote archives into a database and then displaying the data online for further analysis. This presented several challenges specific to historical humanities projects:

  • Locations no longer exist or have changed: ports/countries/regions/provinces stated in the source material are now different.
  • Uncertainty of shipping information: Source material can contain information that was expected to happen, but may not have actually occurred.
  • Ambiguous event locations: the details of storms or attacks by pirates are based on personal accounts rather than GPS.
  • Date normalisation: the different date of adoption of the Gregorian calendar across Europe means that a date in one location could be a different date in another. Further exacerbated by the ‘type’ of calendar used (E.g. Nativity) and therefore the month/day that the year itself rolls over to the next causing differences between certain countries.
  • Date uncertainty: An event or arrival/departure at a port could be ‘around’ a certain date but not definitely on it.
  • Differing currencies and silver equivalence: Many different currencies are used, each with their own denominations and a different weight of silver per coin.
  • Metrology – weights and measures: sacks, bags, crates and barrels, all affect the comparison of cargo.
  • Different spellings of entities: the names of the same port, person and vessel can be recorded differently across cases.
  • No WiFi: the archives are often in older government buildings or museums.

The decided solution was to provide each researcher with an offline database to capture the source material, and then regularly upload the material to the main online database from where the data can be displayed on the project website.

The database is fully relational, allowing data to be searched across all fields and tables (vessels, legs, events, averages) and displayed as both tabular datasets and using geo-mapping software.  

(Ian Wellaway)

Specific Issues:

Here below some short descriptions of the material being uploaded onto the Database, and of the peculiar challenges of each set of data.

Antonio is uploading the data relative to average procedures and reports declared in Genoa from 1590 to the end of the seventeenth century. Giuseppe Felloni’s paper cards have been used as the original model for the Database data capture. Antonio started by improving the format of these cards to make sure that every detail in the sources is appropriately recorded. Just to give an example, the original cards did not record weather events, and these are instead crucial for the project. After a few months of improving the data entry format to record every relevant detail in the sources, and checking the paper cards’ data against the original documents, together with Jake Dyble, Antonio developed a protocol for recording the documents in the offline database. This was then reviewed by all the team members to make sure it is fit for purpose for all the different data in the regions under investigation.

The data from the Consoli del Mare of Pisa is very rich, since every case, with a few anomalous exceptions, contains a calcolo as well as the narrative of the voyage. While these calcoli are not as detailed as the Genoese cases – they generally only give one weight for the cargo, they provide sparse detail on packaging, they do not generally include local values of cargo – this means nevertheless that each case contains considerable qualitative and quantitative data. The particularities of the Livornese data have prompted at least one significant change to the database. The Consuls of the Sea often reduced the amount awarded in average and the calcolo lists only this reduced total as lump sum, rather than as itemised damages. Itemised damages are instead recorded in documents and receipts submitted by the captain according to his own estimates. The existing database architecture could not have captured this data without considerable confusion, hence we developed a distinction a new ‘damages claimed’ tab, where this could be recorded. In another case, the presence of multiple averages – GAs and PAs – within the same voyage prompted the development of a tab whereby damages might be assigned to differing risks (i.e. ship, freight, cargo, and crew) depending on the average in question. My own most significant contribution to the architecture of the database has been the establishment of a series of protocols in conjunction with Antonio Iodice which ensure, as far as possible, a uniformity of styles in entering data, both in the lifetime of the project and afterwards. I have also been instrumental in opening up the database to third party collaboration with Unitre, creating a paper form which could be used to capture entries without a computer, advising Ian on the design of a custom database for Unitre users, and writing a series of clear instructions for electronic data capture.

My data derives from the registers of the Chambre (1668-86) and Compagnie générale des assurances et grosses aventures (1686-c. 1710). This has posed unique challenges, as these were institutions that dealt with general average from the vantage point of marine insurance. Consequently, the basic unit of study in my case has been the insurance policy rather than the GA declaration or the calculus. In the case of the Chambre, there are approximately 6000 policies, while there are approximately 4000 for the Compagnie. Based on these policies, we can track the subsequent events of numerous insured voyages: declarations of average and abandonment reveal the damages to, and losses of, insured vessels and merchandise at sea. Working with Ian Wellaway, the database has been tailored to facilitate the analysis of the valuable underwriting data provided by the registers. In particular, users wishing to study the nature of pre-actuarial underwriting will be able to use the database to track the appetites, specialisations and ultimate track records of specific underwriters over time.

My contribution to the project is the analysis of general average within the context of the Carrera de Indias in the sixteenth and seventeenth century. The relative documents are kept in the Archivo General de Indias, in Seville. I am concentrating on those cases where a general average procedure was at the basis of litigation, and so I am working on a wide range of judicial proceedings, presented before the court by the maestres of the ships that sailed the oceanic voyages, against the interested parties in the cargo of the ship. Considering the particularities of the Carrera, and its convoy-based navigation, the data needs to be seen within this complex contest. I am uploading to the database the following: the identification of the parties (ships, masters, merchants), dates of the voyage, cargo type, legs of the voyage, navigation in fleets, command of the fleet, and, if appropriate, the evidence related to the ruling declaring the general average.

The Sea Protests preserved in the Maltese Notarial Archive range from 1812 to 1879, a crucial transitional time in the evolution of the concept of ‘general average’. They represent an exceptionally valuable type of evidence from the documentary point of view as ‘sea protests’ within the British world were handled by notaries, and therefore very little has survived, and provide a great portrait of English trade in the nineteenth century Mediterranean. The data provided is the same as in other earlier types: nationality of the vessel, master, crew, nature and value of the cargo,  ports of origin, mainly British or from the Black or Azov Sea, such as Russians, Ukrainians and Romanian ports; rarely there are instances from across the Atlantic, with North or South American ports such as Boston or Rio de Janeiro. The vessels are mainly British, but also of other European nations such as Scandinavian, German, Flemish, Dutch, Italian. The master’s city of origin usually matches the harbour in which the vessel is registered. The protests provide details about weather conditions, sometimes as accurate as daily reports on the direction of the winds. In a few instances, there are reports on the health, both mental and physical, of the master or members of the crew (from ‘disordered innervation caused by mental excitement’ to signs of ‘mental derangement’).

Public presentation:

The Database, and its potential for economic history at large, was a focus of an interactive session at the postponed European Social Sciences History Conference 2021 (Leiden, March 2021) in the panel: Representing Risk in Seventeenth-Century Europe: Maritime Averages, GIS and the Digital Humanities.


We are in the process of producing a critical multilingual glossary of the legal and economic terminology employed in GA documentation. This will be freely available from this website in the coming months.