Gerd Hagmeyer-Gaverus was IT and Programme Director at SIPRI from 1984 to 2015.
In the early 1990s, a small peace research institute in Stockholm became one of the pioneers of what the world would come to call the internet—one donated server, one negotiated membership, one reluctant researcher at a time.
First steps into the digital world
I joined SIPRI as a Researcher in 1984, following a period at two universities in Germany and at Virginia Tech in the United States. Particularly based on the US experience, I brought with me something most of my new colleagues did not yet have: a clear sense that technological developments were about to utterly change the worlds of publishing and communications.
SIPRI’s mission to gather, analyse and disseminate information on armaments, disarmament and international security was one that depended entirely on the free flow of information across borders. The Institute had been publishing regularly on questions of disarmament and security since its founding in 1966. Its most important vehicle was the SIPRI Yearbook, produced in partnership with Oxford University Press (OUP), one of the world's great academic publishers. And yet the process by which that partnership operated was, by any modern measure, astonishingly slow.
The manuscript would first be typed at SIPRI, then carefully proofread by the authors and editors. When the text was finally deemed ready, the manuscript was airmailed from Stockholm to Oxford. At OUP, it was typeset, a process that introduced its own errors and required its own corrections. The typeset proofs were then mailed back to Stockholm. Corrections were marked and the proofs returned to Oxford. This cycle repeated itself, back and forth across the North Sea, approximately three times before a manuscript was finally ready to print.
When I arrived at SIPRI, one of my first suggestions was simple: don’t mail the manuscript at all. A modem could transmit the text to Oxford directly, eliminating weeks from the process.
Later, with the arrival of WYSIWYG word processing, we went further still. SIPRI began producing camera-ready copy in-house: finished, typeset pages prepared at our desks in Stockholm and sent straight to the printer. The publisher's typesetting stage was eliminated entirely. Weeks became days.
Three addresses and a modem
The same logic animated everything we would later do with advances in information and communications technology. Each reduction in the distance between writing and reading, between research and dissemination, helped the fundamental mission of the Institute. Any increase in the speed and reach of communication would serve SIPRI's purpose. The question was never whether to adopt new technology; it was how quickly we could do it.
For example, in around 1990 SIPRI joined a network called the Association for Progressive Communications (APC), one of the earliest international civilian email networks on the planet. The process for sending messages was manual and the whole system required a patience that most people simply did not have. But it worked and, even if the speeds were glacial, it was far quicker and more reliable than the post.
I remember vividly the day I sent an email to Moscow, where a SIPRI delegation was meeting. The email had to be rerouted through San Francisco—physically, across the Atlantic and back—before it arrived at its destination. It took a couple of hours. At the time, that felt nothing short of miraculous.
In those early years, SIPRI had three email addresses. They belonged to the researchers who were most curious about the new technology, the ones willing to sit beside a humming modem and wait. I was among them.
Yet SIPRI was (and is) an international institution, constantly welcoming new researchers, with many arriving from the USA where email was already becoming routine. So the pressure for new accounts grew. And every new account came with a cost. I fought quietly for each new address, for each new connection. I knew, with a certainty I could not always explain, that this was the future.
A revelation at KTH
The turning point came in 1993, when I was shown something new at the Royal Institute of Technology (KTH) in Stockholm: the internet. The demonstration introduced me to something called a URL, a string of characters that could connect your computer, through a telephone line, to a server on another continent. You could read a paper published at an American university. With the help of a new piece of software, a browser called Mosaic, you could search a library database in England. You could jump, link by link, from one institution to another, following threads of knowledge across the globe in minutes. The collections of these links, curated by universities, libraries and research institutes, were the early navigational tools of the web.
To appreciate what that moment meant, one must understand the scale of what existed. By the end of 1993, the entire World Wide Web comprised perhaps 2000–3000 websites. Sweden, despite being an unusually early adopter, likely had just a handful of web presences. To be on the internet in Sweden at that time was not to join a crowd. It was to help build one.
I understood that SIPRI would not only benefit from going online, it had a duty to do so. Convincing my colleagues was another matter. The research community at SIPRI was, at that time, deeply sceptical of publishing anything on the web. Our publications—the SIPRI Yearbook, the research papers, the data—were our currency. To give them away freely on the internet seemed, to many, like institutional suicide.
When my attempts at persuading SIPRI to acquire this new technology failed, I decided to act on my own initiative. I turned to the major technology companies of the era, the firms whose hardware and software would form the backbone of any serious internet presence. Their response exceeded anything I had dared to hope for. Cisco provided a router. Sun Microsystems provided a server. Oracle provided a database system. Together, the value of these donations exceeded 500 000 Swedish kronor.
To be on the internet, SIPRI also needed a connection to it. The commercial option, the Swedish telecommunications monopoly, was far too expensive. The new Swedish University Computer Network (SUNET) was affordable, but SUNET's membership criteria were written for universities and their faculties, not independent research institutes. I applied anyway, arguing that we were, in all but name, part of the academic community that SUNET existed to serve. After careful negotiations, we were accepted.
Rickard Schoultz of KTH, one of the true pioneers of the Swedish internet, installed our new server. SIPRI also became one of only three Swedish institutions invited to join the Sun Microsystems SunSITE programme—a network of servers designed to support educational and research institutions with communications technology.
And by 1994, SIPRI was part of the infrastructure of the Swedish internet, and our website was among perhaps 50–150 in the whole country.
Populating the SIPRI website
The new SIPRI website, which I then programmed, was not merely an institutional brochure. From the very beginning, the ambition was to make our data publicly accessible. Two flagship datasets, the Military Expenditure Database and the Arms Transfers Database, were the first to go online, using the database system donated by Oracle. Some of the earliest data was also published as downloadable spreadsheets, because interactive web interfaces were not yet widely understood. Over the succeeding decades, the portfolio of SIPRI’s online databases would grow and evolve, reflecting the Institute’s changing research programme as well as the possibilities offered by new technologies.
The decision to make SIPRI’s data freely available was not without controversy and internal resistance. Some researchers worried about misuse or misinterpretation of the data if users did not read the accompanying notes, methodological explanations and source chapters. Others believed that SIPRI could generate significant revenue by restricting access to paying subscribers. The counterargument, which ultimately prevailed, was straightforward: the subscription income would never cover the costs involved and SIPRI's mandate was to provide information, not to sell it. That decision has shaped the open-access culture of SIPRI's web presence ever since.
The Institute’s relationship with OUP meant that, until the mid 2000s at least, most of our publications could not be made available in full via the SIPRI website. However, we made some shorter resources available for download as well as a condensed version of the Yearbook, translated into several major languages.
As time went on, long-form SIPRI research publications—with the exception of the Yearbook—became rarer. A range of new publication formats began to appear, self-published by SIPRI and made freely available for download. As online video, sophisticated data visualization tools and social media became available, SIPRI was quick to take advantage of these, allowing an ever-growing global audience to benefit from the Institute’s world-leading research.
A reflection on SIPRI at 60
SIPRI celebrates its 60th anniversary this year. Looking back, I was part of SIPRI’s story for half of that time. When I joined SIPRI in my early thirties, I planned to stay a couple of years at most. I was a social scientist and I wanted to work on international security and peace research. SIPRI was the natural place to do it. I ended up staying for more than 30 years, until I retired in 2015. The work and the environment at SIPRI stayed interesting and stimulating the whole time, and we believed that our work could change things for the better.
SIPRI was (and remains) a truly international workplace. Researchers came from all over the world, often for a few years but sometimes just for a few months. As a result, the Institute kept gaining new perspectives. The work also involved international networking: conferences with the global community of security and disarmament researchers, and meetings at the United Nations, the European Union and many regional organizations. Along the way, we met senior government officials and other prominent figures: Soviet President Mikhail Gorbachev, the Dalai Lama, US media magnate Ted Turner, and many more.
I would not have wanted to miss any of it. It was never just a nine-to-five job. It was a commitment to working for a better world. As SIPRI enters its seventh decade, independent research on armaments, conflict and peace is as necessary as it was in 1966. I am grateful to have been part of that story.