Rail advocates often call me “anti-transit,” probably because it is easier to call people names than to answer rational arguments. I’ve always responded that I’m just against wasteful transit. But looking at the finances and ridership of transit systems around the country, it’s hard not to conclude that all government transit is wasteful transit.

Nationally, after adjusting for inflation, the APTA transit fact book shows that annual taxpayer subsidies to transit operations have grown from $1.6 billion in 1970 to $24.0 billion in 2012, yet per capita ridership among America’s urban residents has declined from 49 to 44 trips per year. A lot of that money ends up going to unionized transit workers, but the scary thing is that these workers have some of the best pension and health care plans in the world that are mostly unfunded–which means that transit subsidies will have to increase in the future even if no one rides it at all.

Capital and maintenance subsidies are nearly as great as operating subsidies, largely due to the industry’s fascination with costly rail transit. In 2012, while taxpayers spent $24 billion subsidizing transit operations, they also spent nearly $10 billion on maintenance, and more than $7 billion on capital improvements. In 2012, 25 percent of operating subsidies went to rail transit, but 56 percent of maintenance and 90 percent of capital improvements were spent on rails.

Who, other than rail contractors, union members, and other transit agency employees, is enjoying the benefits of all of these subsidies? To answer this question, I went to the Census Bureau’s American Community Survey page and downloaded table B08519, which shows how people get to work by income class, for states and metropolitan areas.

Nationwide, the data reveal, 5.0 percent of workers take transit to work. For low-income workers–those making less than $25,000 a year–the share is only slightly higher, at 5.9 percent. The shares are lower for other income classes except for people earning $75,000 a year or more, 6.1 percent of whom take transit to work. Where just under 1.2 million people who earn less than $10,000 take transit, more than 1.3 million people who earn more than $75,000 take transit.

About half of those high-earning transit commuters live in the New York metropolitan area. But this isn’t too surprising, as 40 percent of all American transit commuters live in the New York metro area. The high-income workers who take the most advantage of transit subsidies, relative to their low-income counterparts, seem to be mainly in the suburbs of New York, Chicago, Seattle, Washington DC, and–curiously–Idaho Falls.

Even in most places where low-income transit riders outnumber those who earn more than $75,000 a year, transit usage is not heavily weighted to the poor. In the Portland metro area, about 8.5 percent of low-income workers take transit to work compared with 6.1 percent of all workers. In the San Francisco Bay Area, it is 18.5 percent vs. 16.1 percent. In Tampa-St. Petersburg, it is 2.6 percent vs. 1.2 percent.

I’ve recently looked at data for Pinellas County, Florida, whose transit agency wants to build a $1.5 billion light-rail line. The county has more than 400,000 workers, but only 6,000 take transit to work. Of about 14,000 commuters who live in households with no vehicle, 41 percent drive alone (presumably in borrowed or employer-supplied vehicles), 12.5 percent carpool, and just 15 percent take transit to work.

This is not particularly unusual. In nearly 75 percent of the nation’s 315 largest urban areas, more commuters who live in households that lack a vehicle nevertheless drive alone to work than take transit. Nationwide, 21 percent of commuters who lack an automobile drive alone to work, but in some urban areas it is much higher: 51 percent in Riverside-San Bernardino, 42 percent in Raleigh, 40 percent in Kansas City and San Jose, 39 percent in San Diego, 36 percent in Indianapolis, and 31 percent in Tampa-St. Petersburg to name a few.

Most people believe we originally decided to have government take over transit to help low-income people who were transit-dependent. In fact, Congress first passed the Urban Mass Transit Act of 1964 to rescue commuter trains in New York, Chicago, Philadelphia, Boston, and San Francisco, whose private operators wanted to terminate service. In other words, transit was really a subsidy to big-city downtown property owners, not low-income workers.

Even if you want to help low-income workers, there doesn’t seem to be many left who are truly dependent on transit anymore. Between high rates of auto ownership even among low-income people, the growing use of shared rides, and the soon-to-arrive self-driving car, there doesn’t seem to be much use for transit anymore.

I’ve noted before that urban planning professor David King worries that transit is “not living up to its social contract” of providing “the social benefits of mobility for non-drivers.” This view is affirmed by my friend Wendell Cox in New Geography, not to mention this graph comparing subsidies to different Bay Area transit systems and the race of the patrons of each of those systems.

One reason transit agencies no longer worry about low-income people who lack automobiles is that there aren’t very many of them left. Most non-vehicle households may be that way by choice, not by circumstance. After all, more than 20 percent of no-vehicle households are in New York, a state that has only 6 percent of the nation’s population.

Unfortunately, the census tables don’t show vehicle ownership by income, but even to the extent that some low-income households lack cars, it would cost a lot less to give each one a car than to continue subsidizing transit at the rate we do. Considering that less than 5 percent of the households in many states lack autos, I estimate that no more, and probably far less, than 5 million low-income households are without a motor vehicle. Less than two years’ worth of transit subsidies could buy Nissan Versas or similar cars for all of those people.

So transit isn’t for low-income people anymore. We also know that transit isn’t particularly green: except in New York and a few other big cities, transit uses a lot more energy and emits a lot more pollution, per passenger mile, than driving. Nor, outside of New York, does transit carry enough people to relieve much congestion, especially when you consider that the real congestion problem is poorly priced roads, a problem that isn’t solved by providing transit alternatives.

All of which leads me to conclude that there is no longer any sound reason for giving $41 billion in subsidies to transit each year. Transit has become nothing more than a source of political favors to unions, downtown property owners, and rail contractors. As self-driving cars will soon begin to provide mobility for the disabled and those too young or old to drive, it is time for regional planning agencies to stop building obsolete rail transit lines and start thinking about phasing out public transit subsidies over the next two decades.