It has been argued that earmarks can’t have any effect on the level of goverment spending because earmarks only specify how already-budgeted spending will be allocated.  Earmarks don’t represent additional funding, they simply dictate how an agency spends its funds.

But this assumes that the ability to earmark spending doesn’t change the incentives to spend in the first place. Here are two arguments why they must.

  1. Earmarks raise spending. Suppose you are deciding on your grocery budget but your wife is going to do the shopping.  Whatever you budget, she is going to spend it to maximize her own utility function which is not the same as yours.  Your marginal return for every dollar spent is smaller than if you were doing the spending.  Your grocery budget is determined by the condition that the marginal return on groceries is equal to the marginal return on Ahmad Jamal concerts at the Regatta Bar in Cambridge (last week:  awesome.)  This happens at a lower level of spending when your wife has discretion than when you do.  If you could “earmark” the grocery spending you would get exactly what you would buy if it were you doing the shopping.  So the earmarks raise spending.
  2. Earmarks Reduce Spending. Legislation is not decided by a single agent who is maximizing a single utility function.  Legislators have to be bribed to get their support.  Suppose that a crucial supporter wants a road built in his state.  With an earmark this can be achieved directly.  Without an earmark, you have to appropriate general infrastructure spending and hope that he gets treated well by the agency. For the same reason as above, the marginal value to your colleague of a dollar allocated to general infrastructure is lower than if the spending were earmarked.  So you have to increase the budget in order to achieve the bribe he requires.