The federal government’s budget deficits are pushing the nation toward a fiscal meltdown, yet our leaders can’t seem to curb their zeal for infrastructure spending. President Obama has been pushing a $50 billion package for infrastructure and will likely include a similar plan in his upcoming budget. For their part, most Republicans eagerly pursue all the spending they can get for road, rail, airport, and dam projects in their districts.

The enthusiasm for infrastructure spending is emblematic of broader problems in the bloated federal budget. Federal politicians love to intervene in activities that should be left to state governments and the private sector. And they claim that new and expanded programs will produce great results even after decades of waste on existing policies.

A report by Obama’s Treasury Department justifying his infrastructure plan called for “well-targeted” and “carefully selected” investments that are “directed to the most effective programs.” But federal infrastructure spending has never worked that way. For more than a century, it has been shot through with politics, lobbying, and bureaucratic fumbling.

Consider the two oldest federal infrastructure agencies — the Army Corps of Engineers and the Bureau of Reclamation. Their spending, often on water infrastructure, has always been based on pork-barrel politics, not “directed to the most effective” projects. Both agencies have been known for fudging their analyses of dubious proposed projects to gain approval. In his 1971 book on the Corps, distinguished engineer Arthur Morgan found that “there have been over the past 100 years consistent and disastrous failures by the Corps in public works areas … result[ing] in enormous and unnecessary costs to ecology [and] the taxpayer.”

Army Corps and Reclamation projects have often been aimed at benefiting private interests, not the public interest. For example, the Army Corps’ costly Mississippi River–Gulf Outlet Canal near New Orleans, which was completed to aid the shipping industry, failed to benefit the regional economy. It also ended up harming the general public interest by magnifying the damage caused by Hurricane Katrina.

As to how “well targeted” federal infrastructure spending actually is, look at Amtrak. Its investment is dispersed to low-population areas where passenger rail makes no sense. Routes through rural areas account for only a small fraction of riders and cause most of Amtrak’s financial losses. Those routes exist because every lawmaker wants a train through his or her state, but it means that investment gets steered away from where it is really needed, such as the Northeast corridor.

It’s a similar story with federal highway spending. Some states, such as Texas, have growing populations and therefore greater needs. But comparison of the federal highway funding they receive with the federal taxes they pay shows they consistently get the short end of the stick.

Mismanagement is endemic in federal infrastructure projects. Highway projects, energy projects, airport projects, and air-traffic-control projects often end up costing far more than promised. Most infamously, the federally backed Big Dig highway project in Boston eventually cost $15 billion, five times more than the original estimate. But this is nothing new. As far back as 1836, investigations by the House Ways and Means Committee found that dozens of Army Corps infrastructure projects had gone over budget.

The worst aspect of federal infrastructure funding is Washington’s habit of replicating its mistakes across the nation. The building of high-rise public-housing projects, for example, was a horrible idea with disastrous consequences, and federal funding helped spread it nationwide during the 20th century. Or consider today’s federal subsidies for light-rail projects. They have biased cities across the country to approve these expensive projects, even though they are less efficient and flexible than bus systems.

When the federal government allocates funds for particular types of infrastructure, the states are eager to grab their share, discounting any concerns about long-term efficiency. For light rail, cities typically receive federal subsidies for up-front capital costs, but down the road these systems have higher operating and maintenance costs than do buses.

High-speed rail represents another federal effort to create a one-size-fits-all solution for the country. The economic justification for high-speed rail is weak, yet the Obama administration is trying to impose its grand rail vision on the whole nation. Such infrastructure decisions should be left to the states. If California wants to blow its own money on a boondoogle rail system, it can do so, but the federal government shouldn’t foist the costs on the rest of us.

The good news is that there is no need for federal intervention in most infrastructure. In 2012, federal infrastructure spending included $42 billion for highways, $20 billion for water and power projects, $14 billion for urban transit, $12 billion for community development, $6 billion for housing, and $4 billion for airports. Those activities are all state, local, or private in nature. Highways and water infrastructure should be transferred to the states. Urban transit, housing, and community development should be left to the cities. And airports should be privatized, as they have been in many other countries.

There are few, if any, advantages to funding infrastructure at the federal level, and long experience shows that there are many disadvantages. Rather than increasing federal infrastructure spending and adding to the federal deficit, policymakers should sort through current programs and start handing back responsibilities to the states and the private sector.