The Development of Open Access Journal Publishing from 1993 to 2009

Mikael Laakso, Patrik Welling, Helena Bukvova, Linus Nyman, Bo‐Christer Björk et al. (+1)•View original
OverviewAccessibleshimmer voice
Picture this: you're on a bus, scrolling through your phone, trying to read the latest study about a treatment a friend is considering. You find the article. You hit a paywall. Twenty, forty, a hundred dollars for a single paper. Most of us stop there. That wall doesn't just block curiosity. It narrows who can learn, who can verify, and who can build on what we've already discovered. Open Access is the idea of taking the lock off that door. It means the full text is free to read online, and the costs of publishing are covered in other ways. That simple shift changes who gets to participate in science: students, journalists, clinicians in small hospitals, entrepreneurs with an idea, and yes, you on that bus. Let's get concrete. There are two main ways research becomes free to read. One is called Gold Open Access. Think of a Gold Open Access journal as a magazine that gives all its content away online, from day one. The money comes from somewhere else—sometimes from a university or research funder, sometimes from a fee paid by the author's grant. Within Gold Open Access, there are flavors. Direct Open Access is the straightforward one: the whole journal is open, no strings attached. It accounts for roughly six out of every ten Gold Open Access cases. Delayed Open Access opens the door later—after an embargo—so you wait months before the article becomes free. That's about one in seven. And Hybrid Open Access sits inside traditional subscription journals; an author can pay to make just their article free, while the rest of the journal stays behind a paywall. That's roughly a quarter of Gold Open Access. The other path is called Green Open Access. Here, the journal may still be paywalled, but the author shares a copy—often the peer-reviewed manuscript—on a personal site or an institutional or subject repository. If you've ever landed on a free version via a university page or a site like ArXiv or PubMed Central, you've benefited from Green Open Access. In 2008, about twelve percent of all journal articles had a Green Open Access version available. For a reader, that means a second way to get in, even if the publisher's gate is locked. Why does this matter beyond the warm glow of sharing? Because it changes the flow of attention. Search engines now surface multiple versions of the same paper. Google Scholar will often list the publisher's copy right next to a free version. You don't have to know which repository to check. You just click the top link that doesn't ask for your credit card. That discovery shortcut increases the odds that a paper will be read and cited across fields, across languages, and outside of academia. It pushes science from private reading rooms into public conversation. But for years, the big question—how big is Open Access, really?—lacked a crisp answer. As Ware and Mabe pointed out, the field had been mapped in pieces. Researchers took different snapshots of different corners. Crawford looked at the late 1990s, but with a narrow, mortality-focused lens. Wells and Hedlund hunted early pioneers, and their lists were incomplete. Morris examined thousands of journals in the Directory of Open Access Journals, but didn't tease apart which journals were born open and which had converted later, or rebuild year-by-year volume. Each of those efforts was valuable. None of them gave decision-makers a stable, longitudinal view. If you were a funder or a library trying to plan, the numbers were a moving target. That's where a study by Mikael Björk, Annikki Roos, and Mari Lauri stepped in. Their ambition was simple to say and hard to do: build a trustworthy map of Open Access journal publishing from 1993 through 2009. No complete index of Open Access articles existed. So they defined the population with care, and then counted what actually happened, year by year. They chose the Directory of Open Access Journals as their frame, because it had the clearest inclusion rules and the widest Open Access coverage. The Directory of Open Access Journals is not perfect. It lists journals more fully than it lists individual articles, and it removes ceased titles, which can hide some history. But it was the best available starting point. The next step was the heavy lift. They decided not to rely on big commercial indexes like the Institute for Scientific Information or Scopus for article counts. Only a small share of journals in the Directory of Open Access Journals has usable article-level data there, and those databases skew toward established, often Western publications. A survey of journal editors sounds efficient, but response rates would be uneven, and the contact details weren't reliable. So they went manual. A multilingual team visited journal websites and captured, year by year, how many articles were published. They built a custom data-entry tool so everyone counted the same way, and so every entry could be audited later. It's the research equivalent of walking the shelves yourself instead of trusting that the catalog is up to date. Because Open Access journals vary wildly in size and reach, they split the population and sampled accordingly. One stratum contained big journals—more than two hundred articles a year, and visible in the major indexes. All forty-four of those were included. The second stratum captured the long tail: small and midsize journals. From that group, they drew a random sample of 519. They added a third, separate "pioneer" sample of 304 journals known to be active before 2000, to make sure the early story wasn't washed out by later growth. And they corrected the record where needed. In almost a third of the sampled journals, the "start year" in the Directory of Open Access Journals recorded when the title became open access, not when it first began publishing. They went back and fixed 147 such cases, restoring the real beginnings. When they turned counts into estimates, they weighted the results to reflect each stratum's size. That way, a handful of giant titles wouldn't drown out hundreds of smaller ones. They also did a separate cross-check to see how Open Access spread within the universes of Ulrich's, the Institute for Scientific Information, and Scopus. There were limitations. If the Directory of Open Access Journals drops a ceased journal, it's harder to reconstruct that journal's arc. Some journals switched business models along the way. Others disappeared. Even so, the approach brought consistency to a landscape that had lacked it. Here's what the numbers say. Open Access sprinted. In 2000, there were about 740 Open Access journals, publishing roughly 19,500 articles. By 2009, the count had climbed to about 4,769 journals and 191,850 articles. That's not just more journals. The average number of articles per Open Access journal also rose, toward forty per year. Put those figures against the broader backdrop Ware and Mabe describe: total scholarly output grows on the order of three to three and a half percent a year. Within that world, the curve for Open Access is steeper. Since 2000, the number of Open Access journals grew about 18 percent annually. Open Access articles grew about 30 percent per year. Nowhere else in scholarly publishing was growth that fast. It helps to think in phases, the way Björk, Roos, and Lauri do. The pioneering years, from 1993 through 1999, were small in absolute numbers but big in momentum. A few brave titles put everything online for free, often run by passionate editors with thin budgets. Then came the innovation years, from 2000 to 2004. Suddenly the infrastructure caught up. Pioneering publishers like BioMed Central introduced business models where authors, backed by grants or institutions, covered costs so readers didn't have to. The Public Library of Science launched high-profile flagship journals that quickly earned respect. Digitization took off, and regional platforms like SciELO in Latin America and J-Stage in Japan gave whole communities a modern stage. Around 2004, mainstream publishers began experimenting with the hybrid model, letting individual authors buy openness for their articles inside traditional journals. At the same time, advocacy swelled, with declarations and mandates pushing the idea that taxpayer-funded research should be freely available. From 2005 to 2009, we enter consolidation. Growth continued, not as a wild percentage spike year over year, but at a strong and steady clip. Open Access publishing volume rose around 20 percent annually through this period. The machine got smoother. The Directory of Open Access Journals cemented itself as the main Open Access index. Free, open-source software like Open Journal Systems lowered the barrier for small journals to set up shop. Creative Commons licenses gave clear, standard ways to spell out how readers could reuse content. And crucially, more high-volume, high-impact titles went open, pulling the average article counts per journal upward. What about durability? Do Open Access journals stick around, or do they vanish once a grant ends? The pioneer sample gives a reality check. Of 175 journals that were born open and active in 2000, 126 were still publishing in 2009. That's a drop of about 28 percent. Of 43 journals that had converted to Open Access by 2000, 36 were still active nine years later, a smaller drop, around 16 percent. And eight percent of pioneers disappeared entirely. There's churn. But there's also resilience. A large majority of early Open Access journals made it through a turbulent decade, even as models evolved and the economy wobbled. Let's zoom back to the models for a moment, because they matter for readers and for budgets. Direct Open Access—the whole journal is free approach—accounts for about 62 percent of Gold Open Access. That's the most visible and straightforward path. Delayed Open Access, at about 14 percent, opens content after an embargo, which can be frustrating if you need it now but still expands the commons in the long run. Hybrid Open Access, at roughly 24 percent, shows that big subscription publishers didn't sit on the sidelines. They added an open door at the article level, paid for by authors or their funders. Alongside that, Green Open Access added another layer of access, with almost twelve percent of all articles in 2008 having a shareable version online, even if the publisher's site charged for entry. Now, imagine being a policymaker or a librarian in 2009, trying to decide where to invest. Before this kind of longitudinal mapping existed, you had anecdotes and scattered tallies. With it, you have a trajectory you can trust. You can see that the share for Open Access reached about 7.7 percent of all peer-reviewed journal articles by 2009, and that the direction of travel was clear. You can plan for author-fee support, prepare repository infrastructure, adjust subscription strategies, and educate authors about licensing. You can do it with numbers that won't collapse when someone asks how you got them. For authors and readers, the practical meaning is just as clear. If you're an author with a grant, you now know there are reputable Open Access venues that publish at scale and are likely to be alive nine years later. If you're a reader outside a university, your odds of finding what you need without hitting a paywall are higher, and getting higher still thanks to search engines that surface free versions. For the public, this is a slow unfurling of a new norm: science you can actually read when it matters to you. It's also a story about measurement. The team's method—define the population with the Directory of Open Access Journals, collect article counts directly from journal sites, stratify by size, add a pioneer check, correct start years, and weight the results—might sound like inside baseball. But it's the difference between a campaign ad and an audited budget. Stratified random sampling means the loud voices don't drown out the chorus. Manual collection means you're not hostage to someone else's blind spots. A custom tool means the team can show their work. And being multilingual and inclusive matters because Open Access has always been global, with strong currents in Latin America, Eastern Europe, and Asia that older indexes sometimes missed. There's another quiet shift here, one that any listener who searches for papers has felt. The rise of Google Scholar and similar tools changed the path from curiosity to the full text. In the old model, you started at the journal's site. If your library didn't subscribe, you hit the wall. In the new model, you start with a question. The search engine gathers versions from the publisher, the author's page, a subject repository, maybe even a national portal. You pick the one that opens. That small design choice amplifies Open Access's value. It knits together Direct, Delayed, Hybrid, and Green into a reader's single, painless experience: click and read. None of this means the work is finished, or that every number is perfect. The team is upfront about their constraints. The Directory of Open Access Journals removes ceased journals, which can blur the start and end of some titles. Some journals hop models. Article-level indexing lags. Still, the study shows you can build a coherent picture without pretending you have a perfect database. You can keep updating it as Open Access evolves, because the method is reproducible and the data are open for others to extend. Here's what matters when you stand back. By 2009, Open Access had grown into a visible slice of global scholarship. It wasn't the majority, but it was no longer a fringe experiment. It had a backbone—thousands of journals, nearly two hundred thousand articles a year, a maturing infrastructure, and a diversity of business models. It was growing far faster than the rest of scholarly publishing. And it was doing what it promised to do: remove access barriers so more people could read, reuse, and build on research. Looking ahead, the path is less about invention than about maintenance and measurement. The same playbook—clear definitions, careful sampling, open data—can extend the map past 2009. As funding mandates spread and more high-profile journals flip to open, the balance between Direct, Hybrid, and Green will keep shifting. The method here gives us a way to see those shifts without guesswork. And the human part will stay the same. People care about Open Access when it intersects with their lives: a diagnosis, a climate report, a local policy debate. The tools that let someone find a free version, the journal models that keep the lights on without charging readers, the policies that cover author fees—these are the levers that make openness real outside the academy. They're not abstract. They're the reason a high school teacher can pull up a study for a class discussion, or a small-town doctor can check the latest meta-analysis on a drug. So if you remember three things from this ride, make them these. Open Access means research you can read without paying, and it comes in two main paths: Gold, where the journal is open, and Green, where authors share a copy. It grew fast between 1993 and 2009—by 2009, about 7.7 percent of all articles were Open Access, spread across roughly 4,769 journals, with nearly 192,000 articles that year—and it did so faster than the rest of scholarly publishing. And the reason we can say that with confidence is because researchers like Björk, Roos, and Lauri did the slow, careful counting, turning scattered snapshots into a map that readers, authors, and policymakers can use. The next time you click a link and the full text just opens, pause for a second. Behind that one smooth moment is a decade-long story of small experiments, hard data work, new business models, and a community deciding that knowledge is better when more people can reach it.

Picture this: you're on a bus, scrolling through your phone, trying to read the latest study about a treatment a friend is considering. You find the article. You hit a paywall.

Twenty, forty, a hundred dollars for a single paper. Most of us stop there. That wall doesn't just block curiosity.

It narrows who can learn, who can verify, and who can build on what we've already discovered. Open Access is the idea of taking the lock off that door. It means the full text is free to read online, and the costs of publishing are covered in other ways.

That simple shift changes who gets to participate in science: students, journalists, clinicians in small hospitals, entrepreneurs with an idea, and yes, you on that bus.

Let's get concrete. There are two main ways research becomes free to read. One is called Gold Open Access.

Think of a Gold Open Access journal as a magazine that gives all its content away online, from day one. The money comes from somewhere else—sometimes from a university or research funder, sometimes from a fee paid by the author's grant. Within Gold Open Access, there are flavors.

Direct Open Access is the straightforward one: the whole journal is open, no strings attached. It accounts for roughly six out of every ten Gold Open Access cases. Delayed Open Access opens the door later—after an embargo—so you wait months before the article becomes free.

That's about one in seven. And Hybrid Open Access sits inside traditional subscription journals; an author can pay to make just their article free, while the rest of the journal stays behind a paywall. That's roughly a quarter of Gold Open Access.

The other path is called Green Open Access. Here, the journal may still be paywalled, but the author shares a copy—often the peer-reviewed manuscript—on a personal site or an institutional or subject repository. If you've ever landed on a free version via a university page or a site like ArXiv or PubMed Central, you've benefited from Green Open Access.

In 2008, about twelve percent of all journal articles had a Green Open Access version available. For a reader, that means a second way to get in, even if the publisher's gate is locked.

Why does this matter beyond the warm glow of sharing? Because it changes the flow of attention. Search engines now surface multiple versions of the same paper.

Google Scholar will often list the publisher's copy right next to a free version. You don't have to know which repository to check. You just click the top link that doesn't ask for your credit card.

That discovery shortcut increases the odds that a paper will be read and cited across fields, across languages, and outside of academia. It pushes science from private reading rooms into public conversation.

But for years, the big question—how big is Open Access, really?—lacked a crisp answer. As Ware and Mabe pointed out, the field had been mapped in pieces. Researchers took different snapshots of different corners.

Crawford looked at the late 1990s, but with a narrow, mortality-focused lens. Wells and Hedlund hunted early pioneers, and their lists were incomplete. Morris examined thousands of journals in the Directory of Open Access Journals, but didn't tease apart which journals were born open and which had converted later, or rebuild year-by-year volume.

Each of those efforts was valuable. None of them gave decision-makers a stable, longitudinal view. If you were a funder or a library trying to plan, the numbers were a moving target.

That's where a study by Mikael Björk, Annikki Roos, and Mari Lauri stepped in. Their ambition was simple to say and hard to do: build a trustworthy map of Open Access journal publishing from 1993 through 2009. No complete index of Open Access articles existed.

So they defined the population with care, and then counted what actually happened, year by year. They chose the Directory of Open Access Journals as their frame, because it had the clearest inclusion rules and the widest Open Access coverage. The Directory of Open Access Journals is not perfect.

It lists journals more fully than it lists individual articles, and it removes ceased titles, which can hide some history. But it was the best available starting point.

The next step was the heavy lift. They decided not to rely on big commercial indexes like the Institute for Scientific Information or Scopus for article counts. Only a small share of journals in the Directory of Open Access Journals has usable article-level data there, and those databases skew toward established, often Western publications.

A survey of journal editors sounds efficient, but response rates would be uneven, and the contact details weren't reliable. So they went manual. A multilingual team visited journal websites and captured, year by year, how many articles were published.

They built a custom data-entry tool so everyone counted the same way, and so every entry could be audited later. It's the research equivalent of walking the shelves yourself instead of trusting that the catalog is up to date.

Because Open Access journals vary wildly in size and reach, they split the population and sampled accordingly. One stratum contained big journals—more than two hundred articles a year, and visible in the major indexes. All forty-four of those were included.

The second stratum captured the long tail: small and midsize journals. From that group, they drew a random sample of 519. They added a third, separate "pioneer" sample of 304 journals known to be active before 2000, to make sure the early story wasn't washed out by later growth.

And they corrected the record where needed. In almost a third of the sampled journals, the "start year" in the Directory of Open Access Journals recorded when the title became open access, not when it first began publishing. They went back and fixed 147 such cases, restoring the real beginnings.

When they turned counts into estimates, they weighted the results to reflect each stratum's size. That way, a handful of giant titles wouldn't drown out hundreds of smaller ones. They also did a separate cross-check to see how Open Access spread within the universes of Ulrich's, the Institute for Scientific Information, and Scopus.

There were limitations. If the Directory of Open Access Journals drops a ceased journal, it's harder to reconstruct that journal's arc. Some journals switched business models along the way.

Others disappeared. Even so, the approach brought consistency to a landscape that had lacked it.

Here's what the numbers say. Open Access sprinted. In 2000, there were about 740 Open Access journals, publishing roughly 19,500 articles.

By 2009, the count had climbed to about 4,769 journals and 191,850 articles. That's not just more journals. The average number of articles per Open Access journal also rose, toward forty per year.

Put those figures against the broader backdrop Ware and Mabe describe: total scholarly output grows on the order of three to three and a half percent a year. Within that world, the curve for Open Access is steeper. Since 2000, the number of Open Access journals grew about 18 percent annually.

Open Access articles grew about 30 percent per year. Nowhere else in scholarly publishing was growth that fast.

It helps to think in phases, the way Björk, Roos, and Lauri do. The pioneering years, from 1993 through 1999, were small in absolute numbers but big in momentum. A few brave titles put everything online for free, often run by passionate editors with thin budgets.

Then came the innovation years, from 2000 to 2004. Suddenly the infrastructure caught up. Pioneering publishers like BioMed Central introduced business models where authors, backed by grants or institutions, covered costs so readers didn't have to.

The Public Library of Science launched high-profile flagship journals that quickly earned respect. Digitization took off, and regional platforms like SciELO in Latin America and J-Stage in Japan gave whole communities a modern stage. Around 2004, mainstream publishers began experimenting with the hybrid model, letting individual authors buy openness for their articles inside traditional journals.

At the same time, advocacy swelled, with declarations and mandates pushing the idea that taxpayer-funded research should be freely available.

From 2005 to 2009, we enter consolidation. Growth continued, not as a wild percentage spike year over year, but at a strong and steady clip. Open Access publishing volume rose around 20 percent annually through this period.

The machine got smoother. The Directory of Open Access Journals cemented itself as the main Open Access index. Free, open-source software like Open Journal Systems lowered the barrier for small journals to set up shop.

Creative Commons licenses gave clear, standard ways to spell out how readers could reuse content. And crucially, more high-volume, high-impact titles went open, pulling the average article counts per journal upward.

What about durability? Do Open Access journals stick around, or do they vanish once a grant ends? The pioneer sample gives a reality check.

Of 175 journals that were born open and active in 2000, 126 were still publishing in 2009. That's a drop of about 28 percent. Of 43 journals that had converted to Open Access by 2000, 36 were still active nine years later, a smaller drop, around 16 percent.

And eight percent of pioneers disappeared entirely. There's churn. But there's also resilience.

A large majority of early Open Access journals made it through a turbulent decade, even as models evolved and the economy wobbled.

Let's zoom back to the models for a moment, because they matter for readers and for budgets. Direct Open Access—the whole journal is free approach—accounts for about 62 percent of Gold Open Access. That's the most visible and straightforward path.

Delayed Open Access, at about 14 percent, opens content after an embargo, which can be frustrating if you need it now but still expands the commons in the long run. Hybrid Open Access, at roughly 24 percent, shows that big subscription publishers didn't sit on the sidelines. They added an open door at the article level, paid for by authors or their funders.

Alongside that, Green Open Access added another layer of access, with almost twelve percent of all articles in 2008 having a shareable version online, even if the publisher's site charged for entry.

Now, imagine being a policymaker or a librarian in 2009, trying to decide where to invest. Before this kind of longitudinal mapping existed, you had anecdotes and scattered tallies. With it, you have a trajectory you can trust.

You can see that the share for Open Access reached about 7.7 percent of all peer-reviewed journal articles by 2009, and that the direction of travel was clear. You can plan for author-fee support, prepare repository infrastructure, adjust subscription strategies, and educate authors about licensing. You can do it with numbers that won't collapse when someone asks how you got them.

For authors and readers, the practical meaning is just as clear. If you're an author with a grant, you now know there are reputable Open Access venues that publish at scale and are likely to be alive nine years later. If you're a reader outside a university, your odds of finding what you need without hitting a paywall are higher, and getting higher still thanks to search engines that surface free versions.

For the public, this is a slow unfurling of a new norm: science you can actually read when it matters to you.

It's also a story about measurement. The team's method—define the population with the Directory of Open Access Journals, collect article counts directly from journal sites, stratify by size, add a pioneer check, correct start years, and weight the results—might sound like inside baseball. But it's the difference between a campaign ad and an audited budget.

Stratified random sampling means the loud voices don't drown out the chorus. Manual collection means you're not hostage to someone else's blind spots. A custom tool means the team can show their work.

And being multilingual and inclusive matters because Open Access has always been global, with strong currents in Latin America, Eastern Europe, and Asia that older indexes sometimes missed.

There's another quiet shift here, one that any listener who searches for papers has felt. The rise of Google Scholar and similar tools changed the path from curiosity to the full text. In the old model, you started at the journal's site.

If your library didn't subscribe, you hit the wall. In the new model, you start with a question. The search engine gathers versions from the publisher, the author's page, a subject repository, maybe even a national portal.

You pick the one that opens. That small design choice amplifies Open Access's value. It knits together Direct, Delayed, Hybrid, and Green into a reader's single, painless experience: click and read.

None of this means the work is finished, or that every number is perfect. The team is upfront about their constraints. The Directory of Open Access Journals removes ceased journals, which can blur the start and end of some titles.

Some journals hop models. Article-level indexing lags. Still, the study shows you can build a coherent picture without pretending you have a perfect database.

You can keep updating it as Open Access evolves, because the method is reproducible and the data are open for others to extend.

Here's what matters when you stand back. By 2009, Open Access had grown into a visible slice of global scholarship. It wasn't the majority, but it was no longer a fringe experiment.

It had a backbone—thousands of journals, nearly two hundred thousand articles a year, a maturing infrastructure, and a diversity of business models. It was growing far faster than the rest of scholarly publishing. And it was doing what it promised to do: remove access barriers so more people could read, reuse, and build on research.

Looking ahead, the path is less about invention than about maintenance and measurement. The same playbook—clear definitions, careful sampling, open data—can extend the map past 2009. As funding mandates spread and more high-profile journals flip to open, the balance between Direct, Hybrid, and Green will keep shifting. The method here gives us a way to see those shifts without guesswork.

And the human part will stay the same. People care about Open Access when it intersects with their lives: a diagnosis, a climate report, a local policy debate. The tools that let someone find a free version, the journal models that keep the lights on without charging readers, the policies that cover author fees—these are the levers that make openness real outside the academy.

They're not abstract. They're the reason a high school teacher can pull up a study for a class discussion, or a small-town doctor can check the latest meta-analysis on a drug.

So if you remember three things from this ride, make them these. Open Access means research you can read without paying, and it comes in two main paths: Gold, where the journal is open, and Green, where authors share a copy. It grew fast between 1993 and 2009—by 2009, about 7.7 percent of all articles were Open Access, spread across roughly 4,769 journals, with nearly 192,000 articles that year—and it did so faster than the rest of scholarly publishing.

And the reason we can say that with confidence is because researchers like Björk, Roos, and Lauri did the slow, careful counting, turning scattered snapshots into a map that readers, authors, and policymakers can use.

The next time you click a link and the full text just opens, pause for a second. Behind that one smooth moment is a decade-long story of small experiments, hard data work, new business models, and a community deciding that knowledge is better when more people can reach it.