More importantly, it probably is not reasonable to hold policy makers and managers (government, business, etc.) to the rigorous standard of causation. Most problems would have to be abandoned against such a threshold. Policy makers and managers are frequently not in a position to wait until there is sufficient evidence to demonstrate causation simply because social and economic phenomena are highly complex. Policy makers and managers are charged with problem solving. To a large extent, policy making and management are far more art than science. Therefore, policy makers and managers typically do not have the luxury of sufficient time to determine whether there is causation. They often have to act quickly, even when information is incomplete and events are in flux. Sometimes, they have just one chance to get things right. If they fail to act, there are bad consequences e.g., opportunity costs. If they make the wrong decision there are also bad consequences. Hence, if there is a reasonably strong statistical relationship between variables--that relationship being correlation--policy makers and managers will often allow such a relationship to inform their judgment. Such a relationship is not a substitute for judgment, but it can guide judgment.
Finally, one other factor deserves mention. Many phenomena are not gaussian in nature. Hence, an orderly normal distribution is not relevant.