Alberto Alesina and Paula Guiliano write,
We have shown how certain institutions need to “fit” with the dominant culture, such as regulatory policies regarding, for instance, the welfare state, financial regulations, or the functioning of the labor market. On the other hand, certain institutions can determine trust and social capital, as is the case for the presence of free city states in Italy in medieval times.
Therefore, statements like “only institutions matter” or “only culture matters” are unnecessarily single-minded and clearly incorrect
Over the years, I have found a number of articles stressing institutions, while other articles stressed culture. I agree with the authors here that it is more complicated.
I think this kind of question clearly calls for different levels of causation, likely divided into ultimate, intermediate and proximate, with the items and ranking of causation to fill in the blanks being genes —> culture —> institutions. In each case, without the former the latter will not arise and be sustained.