Swedes are not known for complaining about rules and regulations. This has been especially visible during Covid-19. The Swedish constitution does not really allow for “real” lockdowns or other attempts at restraining the ability of Swedish citizens to move freely. Therefore, the parliament decided to issue recommendations instead. And many Swedes followed them without too much of a fuzz.
Against this backdrop, a recently published Open Letter addressed at the Swedish parliament should make you curious. The letter, titled “Open Letter to the Swedish members of parliament: The state of ethical vetting requires immediate revision”, utters hefty critique of the current state of ethical vetting. The letter outlines a set of serious problems with ethical review practices in Sweden that especially researchers in the humanities face, and makes four demands.
Background: The Swedish Ethical Review Act
The background for the open letter is the Swedish Ethical Review Act. This is a legal document requiring ethical review for any research that has to do with humans in one way or another. This includes medical and biological research on humans, but also any research that requires working with what is known as “sensitive personal information”. In general the intention of this law is good: Whenever you work with humans, you have to make sure that your research is ethical.
However – and this is the main angle of attack of this Open Letter – the Swedish Ethical Review Act goes too far. Instead of outlining legal boundaries for ethical research, it mandates that every research endeavor that uses human data has to go through a lengthy process of ethical vetting. If you fail to do so, the law imposes hefty fines or even imprisonment. And in that it does not distinguish in levels of seriousness of the researchers’ wrongdoing. So even if you work with entirely public data that is ethical as far as the various guidelines go, failure to request ethical vetting will make you liable to a lawsuit.
The ethical vetting process is excruciating and difficult to navigate. It has to be in Swedish (even if you are not), it is suited for medical, not social scientific research, it requires an analytical strategy before you even had a look at the data, and it is expensive (~$500 per application).
Aside from that, the open letter makes two additional observations.
First, these requirements shift the power balance in favor of the already powerful. Since any research needs to be ethically vetted, and the Ethical Review Authority (Etikprövningsmyndigheten) can simply deny ethical clearance based on any reason, you may end up having your research killed before it even started. In other words, if the authority – a state funded agency – determines you are a little too critical of the parliament, it is well within its rights to deny your request, and prevent you from conducting your research.
Second, this also makes exploratory research impossible: Since you have to provide all variables and methods you wish to use in advance, you cannot tinker with your data to see if there is anything of interest in there. You have to fix what you are going to do in advance. As soon as your research methodology or data shifts, you may have to re-apply for a review – same lengthy process, same costs.
An Example: My Own Ethical Vetting
To make it easier to understand why all of this can be a problem, let me give you an example – my own ethical vetting.
I work in political sociology and look at parliamentary behavior – currently in the U.S. Congress. I am interested in how parliamentarians work, how they behave, and how that behavior shapes policy. For this, I use exclusively public information: I don’t need surveys, data collection or any proprietary data sources. What I need is readily available from the Library of Congress (the parliamentary transcripts) and the U.S. Senate and House respectively (biographical covariates and bill metadata).
Of course, I need to make sure that I conduct my research ethically, but there are enough guidelines out there that I studied in advance to know what I can and cannot do. Even more so, I did my research before to make sure I only use ethically unproblematic data and methods to avoid a more formal vetting process. But alas, the Swedish law doesn’t care: I still had to apply for ethical vetting, and so I did.
My ethical application process took almost an entire year, most of which was dedicated towards reading and properly understanding the requirements for the application, and crafting the proper responses to their questions. The application that I finally submitted had 30 pages and included three abstracts for the same research project with various lengths. In addition, five people were involved in crafting and submitting the application (me, my supervisor, a Swedish native speaker, our department’s head and our institute’s head). And, of course, my institute had to pay ~$500 just to make Etikprövningsmyndigheten have a look at it.
Remember: All of this for using public data that many other researchers have used before me and that are generally understood to be entirely unproblematic. I can’t complain too much, though: Now I know for certain that my research is indeed ethical, and, given the rigorousness of the process, this won’t be easily questioned later on.
Don’t Mix Law and Ethics
This example shows you the main issue with the Swedish Ethical Review Act: It mixes ethical and legal provisions, even though both are incommensurable.
Ethics is fundamentally vague and based on a “case-by-case” approach. Looking at ethnicity, for example, may be ethical in one instance, but highly unethical in another. Furthermore, ethics depends heavily on degree: Examining representatives’ opinions is almost never a problem, but medical interventions on patients require an extensive vetting process.
Law, on the other hand, is made to set out absolute boundaries of behavior in society without all that much wiggle room. In other words, law cannot incorporate the two contradictory statements “Ethnicity is ethical” and “Ethnicity is unethical”. Furthermore, if you do not explicitly write degrees into the legal text, all research is created equal under the law, and thus requires the same process every time.
There are many examples of ethical, but unlawful, and unethical, but lawful research. My research, for example, would be ethical, but unlawful, had I not requested and received ethical clearance beforehand. On the other hand, faking research data would be highly unethical but not directly illegal.
The Swedish Ethical Review Act does not just mandate that people with ethical experience have a look over every research project (which I do agree with), but to a certain degree it also prescribes what counts as ethical.
By baking ethics into law, Sweden effectively froze a single perspective on ethics into law, and that perspective is highly influenced by the ethical requirements of medical research. Medical ethics is much more evolved as a field and much clearer to interpret. But research in the humanities frequently does not impede humans in any way, rendering ethical decisions less clear-cut.
To hammer home this point, consider the General Data Protection Regulation (GDPR) as a comparison: While the GDPR protects many data points connected to individuals (including political views), it explicitly grants an exception for research. In other words, researchers are trusted in doing research with human data properly. The Swedish Law does not include these exceptions — even though it explicitly refers to the GDPR when it comes to defining what counts as “sensitive” data.
What Does the Open Letter Demand?
This brings us to the demands of this letter. Effectively, the letter makes four demands to the members of parliament:
- The Board of Appeals for Ethics Review should not be obliged to drag suspected (minor) violations of the Ethics Review Act in front of a court (i.e.: it is inappropriate to fine or imprison people for failing to request ethical review when the associated risks are minimal)
- Add exceptions to the Ethics Review Act for those research projects where the chance of “significant risk” is very low (effectively this demands the incorporation of the GDPR exceptions into Swedish law)
- Exempt research that uses data that has been “manifestly made public by the research subject” from ethics review (this is another clause from the GDPR that “un-protects” personal data if the subject has explicitly made something public – for example by running for office and speaking as a member of some party in parliament)
- Split up the ethics review requirements into a medical part and a non-medical part (effectively this demands to set up IRBs at the local university level for non-medical research, highlighting the different ethical requirements of medicine vis-à-vis social sciences)
There is a pattern visible in all of these demands. Effectively, they boil down to two things: (a) align the Swedish legislation with the GDPR, and (b) decentralize ethical decision-making.
I am honestly not sure if I can agree with the second issue, since decentralized IRBs can also cause lots of frustration and wrong decisions. The rest of these demands are very reasonable, however, and I personally think that not being subject to criminal charges simply for forgetting to apply for a year-long ethical review process is a good goal to fight for.
What Will Be Next?
My own experience from Germany tells me that open letters generally do not get traction if they aren’t visible in the media. A few colleagues from Sweden have told me, however, that this letter may actually cause movement in the Swedish parliament. At the time of writing it has been signed by more than 2,400 researchers which for a sparsely populated country like Sweden is a high number.
But it remains to be seen if this will actually cause a revision of the current state of Swedish Ethical Research. I personally hope for change, since projects such as mine should not be subject to such a complicated process.
I am convinced that every research should go through a third-party ethical vetting, but the complexity of the vetting process needs to be in proportion to the expected risks. If there are minimal risks involved, such a vetting should take maybe two weeks at most. If there are heavy risks involved, then the process as it is now seems adequate.
Ethics is a difficult field, and getting it right can be tricky. But I agree with the open letter: The system as it is now serves as a deterrent rather than an enabler of research. Ethical review should never discourage researchers or deny a project, but simply ensure that the researchers’ and research subjects’ interests are aligned. I am certain that almost every ethically questionable research project can be made ethical if only you try.