A problem with an infrastructure bill is how we do infrastructure planning: constrained versus "visionary" projects

Transportation planning for large projects is guided by various federal planning requirements.  Projects are planned regionally, through what is called a Metropolitan Planning Organization and guided by what is called a "Constrained Long Range Plan."  

Even though they are for a long period, 20-25 years (and regularly updated), CLRPs are "constrained" by only including projects for which money is highly likely to be obtained.

-- Constrained Long Range Transportation Plan for the Washington Region, MWCOG

That means "pie in the sky" or really important but hard to fund projects aren't in the plan.  

And they won't be able to be funded as part of a federal infrastructure initiative that requires detailed specifics now, unless there is a way to build this possibility into the plan

In transit, such "pie in the sky" projects could be

• connecting North and South Stations in Boston ("3 Reasons Why We Should Build The North-South Rail Link," WBUR-FM/NPR)
• adding back commuter rail lines in Greater Philadelphia, which had been eliminated when they moved to supporting only lines that were electrified, so that they could run underground in Philadelphia
• creating other transit lines in Greater Philadelphia, that have been considered in plans since the 1970s ("Glassboro-Camden light rail proposal takes next step in South Jersey," Philly Voice)
• extending the electrification of passenger trains from Harrisburg to Pittsburgh ("Will there ever be high-speed rail service between Pittsburgh and Philadelphia?," The Incline)
• creating a statewide passenger railroad program in Maryland
• extending the light rail and subway program in Greater Baltimore ("From the Files: Transit Planning in Baltimore County," 2010)
• electrifying Caltrain (which is underway), further extending the MUNI Central Subway Extension project from Chinatown to North Beach in San Francisco
• filling gaps in the Southern California rail transit network and coastal water taxi services in Los Angeles County
• connecting Los Angeles, not just San Bernardino, to Las Vegas by train ("Train Connecting SoCal to Las Vegas Could Be Ready in 2024," Spectrum1 News)
• building maglev between DC/Richmond and Boston
• combining the MARC and Virginia Railway Express lines into an integrated system
• expanding MBTA to Western Massachusetts ("State to study three potential plans for east-west passenger rail," Greenfield Recorder)
• creating a statewide passenger rail program for New York State
• extending the 7 Subway line to New Jersey ("Proposal to extend 7 train into New Jersey revived," NY1 News), creating a subway line in Urban New Jersey along Bergenline Avenue, making the rail services in Greater New York City function like the S-Bahn, merging NJ Transit with LIRR and MTA for through running service, extending the Q Line to LaGuardia Airport, transit service to Staten Island from New Jersey, extending PATH to Grand Central Station, etc.
• water taxi service in Los Angeles County
• funding Michigan's initiative to extend railroad services northward ("Ann Arbor-Traverse City train developers aim for test rides in 2021," MLive)

• creating a Separated Silver Line Metrorail in Virginia and DC, and extensions of the Blue Line to Dale City, the Yellow Line to Fort Belvoir, the Orange Line to Centreville, the Green Line to Laurel and Brandywine
• etc., etc., etc.

And this kind of "constrained planning" approach is true for most categories of infrastructure, such as:

• bicycling and pedestrian facilities and improvements
• water treatment, stormwater and sewer projects (Drinking Water Infrastructure Needs Survey and Assessment, EPA)
• rural and urban broadband
• local bridges--cities in Michigan are selling bridges to private firms which have the money to fix them, and then impose tolls ("The company buying two Bay City bridges wants to connect with the community," Second Wave Media)
• national, state, and local parks ("What is deferred maintenance," National Park Service)
• dams ("Midland failed dams, floods caused $200M in damages to 2,500 buildings," Bridge Michigan)
• public buildings
• the unique needs of San Diego County to support sewage system improvements in Tijuana, Mexico, because overflows pollute local beaches ("EPA says help is on the way for San Diego beaches fouled by sewage from Tijuana," Los Angeles Times)
• etc.

The demands are great but there is no master plan that includes both immediately needed and visionary projects.

And when opportunities to fund such projects come along, like in a once in a couple generation federal infrastructure bill and funding mechanisms, there is no way to get projects into the program because they haven't been planned, with all the necessary design and engineering required to make them "shovel ready."

So they don't get funded.

This is what happened with the American Recovery and Reinvestment Act during the Obama Administration.  The projects were required to be "shovel ready," because they wanted to stoke employment, immediately.

For the most part, only road projects are shovel ready, and as the Brookings Institution points out, such a requirement means that this ends up focusing on repairs ("Eight Years Later: What the Recovery Act Taught Us about Investing in Transportation").

That requirement or bias meant that transformational projects stayed on the drawing board.

 And in terms of high levels of extranormal return on investment from infrastructure projects, it's the extraordinary not the ordinary projects, that tend to build and support long term economic growth.

On the other hand, because of Republican intransigence, and the fear that in 2022 the Democrats could lose control of the Senate, House, or both, the Biden Administration needs to act now, and planning takes a long time.  Waiting for an unconstrained plan could mean losing the opportunity to pass any kind of bill whatsoever ("Biden prepares to move to next phase of his agenda with infrastructure push," NBC News).