I wonder how much of this is correlation vs causation. For instance, if you can afford a couple cups of coffee a day (time and money) perhaps you’re just more well off in general. Coffee might be a bad example in this case because it’s pretty low cost (wines a better example), but my point remains.
My comment was less on the specifics (cost/time) and more on causation vs correlation. See Sop’s comment for a much better potential cause. I agree, the cost/time is a weak argument.