When people talk about “the decline” of Baltimore (which is, in fact, a lovely city), they probably don’t mean it this literally: On Wednesday, a section of East 26th Street collapsed and slid into an embankment onto some train tracks. It’s hard to capture the utter strangeness of watching part of a city block simply fall away, so I recommend you watch this video, about which the Post aptly comments: “Wow, just wow.”

The street collapse—which thankfully caused no injuries—occurred after a nearly 120 year-old retaining wall finally gave way after a long period of heavy rain. The wall, which had protected a stretch of train tracks belonging to CSX, had been the subject of dispute between the company and the city since at least the 1990s, with persistent disagreement over responsibility for its upkeep. In the wake of its collapse, MSNBC’s Steve Benen made the evergreen observation: “Now would [be] an excellent time for the United States to make some investments in transportation infrastructure.” As Benen points out: “Improvements to the nation’s transportation infrastructure are not exactly optional […] It’s not like we’re prepared to stop moving people and products around the country.”

That’s an appealing approach to most policy wonks, but transportation infrastructure spending is not without its skeptics. Just last month, the Brookings Institution’s Clifford Winston challenged the conventional view of transportation infrastructure investment as boring-but-virtuous policy, claiming that “infrastructure costs are inflated by inefficient investments and burdensome regulations.” Moreover, Winston argues, Americans are underpaying for the problems caused by their driving:

Every time Americans use a road, airport, or even cross a bridge, we don’t actually pay for the congestion we cause. That gives policy makers the misleading impression that it is necessary to build new roads, airports, urban transit — and even a high-speed rail system! — to reduce delays. But public officials would likely arrive at a different conclusion if we users were charged congestion tolls, especially during peak travel periods. Demand would fall, transmitting a more accurate signal. Supply and demand would then do its job nicely.

The “myriad of policy inefficiencies” built into the process, Winston argues, makes many of these investments astonishingly inefficient: Over the last decade, for instance, the return on highway investments has been a measly 1 percent.

Incidentally, just one day before the street collapse, the Obama administration announced a new $300 billion transportation bill. And it may not suffer from all the problems outlined by Winston. After all, as Time magazine’s Michael Grunwald notes, the proposal is not exactly your typical highway bill: Instead, it “continues the Administration’s subtle shift towards passenger rail, freight rail, dedicated bus lanes, and other programs that don’t necessarily involve asphalt.” And it contains a provision which might help Americans absorb the full cost of their driving: a “provision allowing states to put tolls on interstates.”

That last item might be popular with Brookings’s Winston, if nobody else. Of course, this is all likely to remain a hypothetical debate for some time: Although infrastructure spending has collapsed since 2008—driven in large part by declines at the state and local levels, the latter of which was at the heart of the Baltimore retaining wall dispute—the House is unlikely to consider new spending on the issue. So if you’re walking down the street after a heavy rainstorm, keep your eyes peeled for any cracks in the pavement. And if you hear that telltale groaning and creaking, don’t waste time finding your way to solid ground.