Consider the Gulf Coast arc from Houston to Tampa, an area growing on the back of the shale energy industry and agricultural exports. The ports of Corpus Christi and Tampa both received federal foreign trade zone status in the early 1980s and have been raising bridges and expanding terminals to prepare for larger ships coming through the Panama Canal — and their modernization also means accelerated export of food, oil and cars from America’s heartland. Their fates are more intertwined than Tampa’s is with Tallahassee or Corpus Christi’s is with Austin, even though they’re in the same state — and yet building out their infrastructure depends largely on the political whims of their respective state capitals. As a result, the region’s ports have built redundant facilities rather than strengthening those best suited to capitalize on new economic connections.

Nor is it just about federal policy. States need to work across borders, too. For example, instead of waging a 1980s Asian-style race to the bottom to attract low-wage auto jobs at Nissan, Honda or Toyota plants, Tennessee and Kentucky should join forces to become an advanced manufacturing hub for the global auto industry, with better cross-border infrastructure. They may end up with fewer plants, but they would be more competitive ones, especially if they could coordinate research and development through the states’ public and private universities.

Where possible, such planning should even jump over international borders. While Detroit’s population has fallen below a million, the Detroit-Windsor region is the largest United States-Canada cross-border area, with nearly six million people (and one of the largest border populations in the world). Both sides are deeply interdependent because of their automobile and steel industries and would benefit from scaling together rather than bickering over who pays for a new bridge between them. Detroit’s destiny seems almost obvious if we are brave enough to build it: a midpoint of the Chicago-Toronto corridor in an emerging North American Union.

TO make these things happen requires thinking beyond states. Washington currently provides minimal support for regional economic efforts and strategies; it needs to go much further, even at the risk of upsetting established federal-state political balances. A national infrastructure bank, if it ever gets off the ground, should have as part of its charter an obligation to ignore state lines when weighing projects to support.

Consider how parts of the Rust Belt could benefit from this approach. A Midwestern high-speed rail network that ran from Southern Illinois to Southern Michigan would not just link wealthy investment hubs like Louisville, Ky., and Columbus, Ohio; by tying in high-unemployment cities like Dayton, it would make it easier for workers to commute to where the jobs are.

Such networks would just as easily help poor and rural areas, like Appalachia. Upgraded transportation corridors between New York, Washington and Atlanta could finally lift Appalachia’s isolated and stagnant towns stretching from New York to Alabama by facilitating investment in farms and vineyards, food processing and eco-tourism.

States will continue to have an important political and regulatory function to fill. But the next president has to move beyond platitudes and implement a serious policy of leveraging new infrastructure investment from home and abroad and backing the shift toward a new urban political economy built around transportation engineering, alternative energy, digital technology and other advanced sectors.

The 21st century will not be a competition over territory, but over connectivity — and only connecting American cities will enable the United States to win the tug of war over global trade volumes, investment flows and supply chains. More than America’s military grand strategy, such an economic master plan would determine if America remained the world’s leading superpower.