Governments and transportation networks are both hierarchically organized. Some state governments finance most of their highways, whereas in other states similar roads are financed locally. Larger governments attain scale economies. However, they also tend to be more bureaucratic and have higher operating costs, all else being equal, because of problems such as span of control. A study was done to relate highway expenditures with the share of expenditure by state governments to determine how governments should allocate expenditure on all roads in a state. Highways are divided into two hierarchical classes (higher and lower), governments into two layers (state and local), and costs into capital and operations and maintenance. Regression models to predict different highway expenditures on each highway class as a function of utilization, capacity, and funding shares are estimated. It is found that there is a share of expenditures by each level of government for each highway class that results in a minimum expenditure for each funding category (capital and operations). That minimum is not very far from typical state and local mixes found in many states. The results can be applied in formulation of efficient network financing arrangements. Policies can be formulated that would help adjust the financial responsibilities of transportation networks between government layers.