Why is it harder for researchers to study social media now?
Because platforms have restricted or withdrawn the access that independent research depended on — and this happened while the questions researchers are asked to answer became more consequential.
What researchers previously had. Application programming interfaces allowing systematic collection of public posts, network structures and engagement data at scale. This underpinned a substantial body of work on misinformation, polarisation, harassment, advertising transparency and public health communication.
What changed:
API access was restricted or priced out of reach. Several platforms introduced charges at levels far beyond academic budgets, or removed free research tiers entirely, and the change was frequently made with little notice.
Research tools were shut down, including transparency tools that had been widely relied upon.
Accounts and projects were restricted, including cases where researchers studying advertising or misinformation had access withdrawn — which drew significant criticism, since studying a platform is not the same as violating its terms.
Terms of service prohibit automated collection, so scraping carries legal and contractual risk even for clearly public data.
Conversation moved to private spaces — messaging apps and closed groups — that cannot be studied by observation at all, regardless of permission.
Why platforms give for the restrictions: privacy and data protection obligations, commercial value of the data, prevention of misuse, and infrastructure cost. Each has some validity, and the privacy argument in particular is genuine and not merely pretextual.
Why the consequences are serious:
Independent verification of platform claims becomes impossible. Only the platform can measure what happens on it, so its own reporting cannot be checked.
Research lags reality, and studies are increasingly based on older data or on the platforms that still permit access — which biases the literature toward whichever platform is most open rather than most important.
Policy is made without evidence, since regulators rely on research that can no longer be conducted.
What is being attempted: regulatory data access provisions, notably under the EU Digital Services Act, which obliges very large platforms to provide vetted researchers with data; independent data donation studies, where users voluntarily contribute their own data; and archival collaborations.
Implementation has been slow and contested.