Whether something is a "real science" is based on whether or not science is done, not the difficulty of the subject matter. Spending so much time trying to narrowly semanticize science in a way that specifically excludes economics is navel gazing. The purpose of these discussions seems to be to discredit economics and discount it's predictive power by defining it as not-science, but predictive power doesn't come from the definition of the word science, it comes from the statistical methods that science uses.
Does it involve research? Does the research produce models with predictive power? Ok, let's use it, just like every policy maker in the world pre-2016.
Does it involve research? Does the research produce models with predictive power? Ok, let's use it, just like every policy maker in the world pre-2016.