Amtrak issued its F.Y. 2016 unaudited financial results last week with a glowing press release claiming a "new ridership record and lowest operating loss ever." Noting that "ticket sales and other revenues" covered 94 percent of Amtrak's operating costs, Amtrak media relations called this "a world-class performance for a passenger carrying railroad." The reality is quite a bit more dismal.

Many new high-tech firms attract investors despite losing money, but a 45-year-old company operating an 80-year-old technology shouldn't really brag about having its "lowest loss ever." The "world-class performance" claim is based on the assumption that passenger trains all over the world lose money, which is far from true: most passenger trains in Britain and Japan make money, partly because they are at least semi-privatized.

Moreover, a close look at the unaudited report reveals that Amtrak left a lot of things out of its press release: passenger miles carried by Amtrak declined; ticket revenues declined; and the average length of trip taken by an Amtrak passenger declined. The main reasons for Amtrak's positive results were an increase in state subsidies (which Amtrak counts as passenger revenue) and a decrease in fuel and other costs.

Ridership grew by 1.3 percent, but passenger miles fell because the average length of trips fell by 3.1 percent. One of the biggest drops in trip lengths was on the New York-Savannah Palmetto. Starting at the beginning of F.Y. 2016, Amtrak added stops at Metropark, New Brunswick, Princeton Junction, and Baltimore-Washington Airport, effectively turning the supposedly long-distance train into a Northeast Corridor train. In 2015, the train's average trip length was 396 miles, but in 2016 that dropped to 257 miles.

A decline in passenger miles means more empty seats. In 2015, Amtrak filled 51.4 percent of its seat-miles; in 2016, this fell to 50.0 percent. In other words, the average Amtrak train is half full; when was the last time you were on a half-full airliner? The biggest declines were on the Washington-Richmond state-supported train, the Seattle-Los Angeles Coast Starlight, and the Auto Train.

Some trains did show an increase in passenger miles. One of the biggest increases was the Chicago-Indianapolis Hoosier State, which saw an 11 percent increase in passenger miles and a 16 percent increase in revenues. This train is supported by Indiana, which got fed up with Amtrak service and contracted it out to another operator, Iowa Pacific. Amtrak is a "partner" because it allows people to make reservations on the train from its web site. But the lesson may be that privatization (or semi-privatization) can result in bigger ridership gains than Amtrak.

The biggest increase in Amtrak's revenue was state subsidies for trains such as Washington-Norfolk, Chicago-St. Louis, Seattle-Eugene, and the California trains. In 2015, these trains earned $1.62 in total revenues for every $1 in actual ticket revenues; in 2016, this grew to $1.76 per dollar. Most of the difference between total revenues and ticket revenues is state subsidies, which grew from $222.9 million to $227.5 million. 

Decreasing costs, not increasing revenues, accounted for most of the increase in the share of operating costs covered by revenues. Fuel costs declined by $53 million. Wages fell by $12 million (though executive salaries grew by $17 million). The biggest savings was a $79 million decline in employee benefits, due to late F.Y. 2015 cuts in both pensions and health benefits.

The focus on the share of operating costs that is covered by revenues conveniently ignores the fact that not all costs are operating costs. Amtrak reported ticket revenues of $2.1 billion and total expenses of $4.2 billion, so passenger fares actually covered just 50 percent of total costs. There's a big difference between 94 percent and 50 percent.

That difference is largely due to an issue that I've noted before, which is that Amtrak has defined away a lot of operating costs by calling them capital costs. It's also difficult to tell how much Amtrak is reducing costs by deferring maintenance on its infrastructure and rolling stock.

The truth is that not much is different from 2015. Amtrak still requires well over a billion dollars in federal subsidies per year. That makes Amtrak a world-class money loser, just like most European state-owned railroads. Despite the implicit promise of "declining operating losses," that's not going to change anytime soon unless Congress kills the program.