The Horse Doesn't Have to Come Before the Cart
I associate the phrase "the horse has to come before the cart" with a common political argument that I think is invalid. Once, I heard someone say that it would be desirable to have less traffic in big cities. However, he didn't like the obvious measures that could be put in place to reduce traffic like congestion taxes or higher prices for parking. He thought that these things should not be done until effective mass transit systems were already in place. Similarly, I once saw someone say that, although she thought abortion was bad, abortion is a consequence of a culture of delayed family formation and (relative) free love. Changing the rules about abortion, her argument went, is undesirable until that culture is changed.
This is my read of the abstract form of the horse-cart argument:
(1) Something (call it X) is commonly done, but undesirable.
(2) The rules could be changed to punish or tax X.
(3) However, this would not change the circumstances beyond the rules that create demand for X.
Therefore, changing the rules to discourage X is undesirable until the circumstances are changed.
When laid out like this, the horse-cart argument seems pretty unconvincing. What I think it misses is that circumstances can change in response to new rules. If the costs of parking are raised, people might carpool. Private businesses might start running bus lines. Citizens might start to vote for mayoral candidates who promise to put in trains. If abortion were made illegal presumably people would price that in and begin to change their decision making. The culture is just the constellation of prevalent beliefs, preferences, and behaviors. Rule changes can obviously have unintended consequences. But the fact that cultural change is one requirement of changing something is by itself no argument against changing the rules. Changing the rules can cause cultural change.
It is true that changing the rules without (somehow) first changing the related cultural scaffolding will impose costs disproportionately on a specific group of people (those who live far from work and need to pay more for parking or move, say). But changing the culture before changing the rules would impose costs on a different group of people (those who now have to live near a loud and annoying train route, for example).
The cost of changing how you act is greater than the difference between the benefit of how you acted previously and how you will act going forward. If the rules change to discourage something that you were previously planning to do, you are likely to be worse off than if the new rules had always been in place. I believe the lawyerly term for this is a "reliance interest." In 1967 Sweden switched from driving on the left (like Britain does) to driving on the right (like nearly every other country does). Although neither way of driving is inherently easier or harder, Swedes struggled to adapt to the change:
If Sweden had started by driving on the (forgive me) correct side of the road, this traffic jam could have been avoided. It is a mistake to count the costs of changing the rules without accounting for reliance interests.
But the horse-cart argument seems to go way beyond what simply recognizing the existence of reliance interests would warrant. And it is (almost by definition) harder to change organic circumstances than rules. So the horse-cart argument seems likely to function in practice as an argument for never changing anything.