Effectiveness research should be given more space in public policy. But not at all costs.
Photo: Einar Nilsen
The previous government extended the opening hours of the Norwegian state-owned alcohol monopoly's sales outlets. However, alcohol consumption did not increase (1) . This means that extended opening hours do not lead to increased consumption, one politician recently claimed on the evening news (1) . This argument is obviously fallacious; the co-variation could be due to anything that might have happened in the meanwhile.
We may soon have a better answer. While planning for the extended opening hours, the Ministry of Health and Care Services contacted the Norwegian Institute of Public Health and requested that they conduct a study (2) . As a result, when the Storting adopted the extension in June 2020, it was introduced in stages in the form of a cluster-randomised controlled trial, developed to be able to identify any changes (3) .
The pandemic showed us that we need better knowledge on the effect of political measures, both during crises and at other times
We need more such studies. The pandemic showed us that we need better knowledge on the effect of political measures, both during crises and at other times. This is no surprise. Suddenly, the whole of Norway was sitting in front of their TVs (there were few other places to be) watching politicians, bureaucrats and scientists discussing school shutdowns and the use of face masks without knowing whether these measures had any real effect on infection rates and without knowing the adverse effects of various types of 'treatment'. In parallel, we were of course waiting for the vaccines to undergo the mandatory testing before they could be rolled out to the population. We could thus quite simply see the discrepancy between the evidence requirement for drug-based interventions on the one hand and political measures with their potentially positive and negative effects on the other.
Just before the summer, two government-appointed expert committees submitted their reports on how to better develop knowledge during crises. One report discussed paving the way for increased use of randomised, controlled trials in public administration (4) . The authors write that one of the reasons why so few such trials were conducted during the pandemic is that the authorities lack knowledge and training in the use of such trials, and that no procedures for this are in place even in normal times.
To rectify this, better cooperation between public administration and researchers is called for. Furthermore, the legislative framework should be reviewed, especially when it comes to the requirement for consent (4) . The regional committees for research ethics should be authorised to waive the requirement for consent when it cannot be met in practice (for example when it must be obtained from all parents of all children in all classes that are included in a research project), given a number of preconditions and thorough prior assessment (4) . In this way, different forms of piloting and testing of interventions will be given more space in public administration than is the case today, and we will be better prepared to implement similar research projects the next time we are hit by a crisis.
Better cooperation between public administration and researchers is called for
This sounds promising, but with some provisos: first, public administration must not fall into the trap of believing that randomised, controlled trials are the answer to everything. This design has many limitations, and such trials are frequently both costly and time-consuming (5, 6) . For example, the results from the study of the effect of extended opening hours are yet to be published. Second, requirements must be set to ensure a distance between those who commission the research (politicians) and those who conduct it (researchers), as is done in drug trials. Third, what is published must be quality-assured through external peer review and disseminated through suitable channels.
Last, but not least, this must be backed by funding. In its report, the expert committee writes that 'the Ministry of Education and Research should also consider strengthening the Research Council's prioritisation of research and innovation projects that include evaluation of interventions using an experimental design ...' (4) . In a situation where grants to basic research are under threat (7) , a shift in priorities away from research and potentially towards poorly designed intervention studies to satisfy politicians who are eager to enact measures is the last thing we need.