Washington finally has a great new (tax-subsidized) convention hotel to go along with its great new (subsidized) downtown convention center. The glass-and-steel building is about as light and inviting as a 1,175-room hotel can be. With the completion of the Marriott Marquis, Washington is now positioned to compete for more and bigger national meetings and conventions.

Unfortunately, in the hypercompetitive world of (subsidized) conventions, there is never a good time to rest on your laurels.

Before long, local boosters will join with the Convention Industrial Complex (CIC) to try to convince us that our billion-dollar investment will all be for naught if we don’t pony up more public funds to expand the existing facilities so that we can go toe to toe with Chicago and Vegas or keep up with the competition from San Diego and Nashville. That’s already happening in Boston, one of Washington’s big rivals, where the Massachusetts legislature has just approved $1.1 billion in additional financing for convention center expansion. And it’s only a matter of time before it happens here again.

To recap the story so far:

It was nearly 15 years ago that ground was broken for a $850 million convention center to replace a smaller, dowdier facility that, in its waning years, managed to generate 350,000 “room nights” for Washington hotels. (Room nights are a common metric for measuring the success of convention centers.) Using the standard CIC playbook, proponents commissioned a consultants’ report predicting that the new center would generate thousands of additional jobs and hundreds of millions of dollars a year in economic benefits, with the entire cost passed on to out-of-towners through a 4.5 percent add-on to the existing hotel tax and a 1 percent add-on to the restaurant and car rental tax. In other words, for Washingtonians, a proverbial free lunch.

As things turned out, rather than the 750,000 room nights a year that the consultants projected, the number has averaged about 500,000. As the accompanying chart indicates, the peak came in 2005 — at 590,000 — and then only because Hurricane Katrina forced groups to move their gatherings out of New Orleans. At the trough of the recent recession, the number fell to about 300,000.

Washington’s experience was hardly unique, as Heywood Sanders, a professor at the University of Texas at San Antonio lays out in a new book, “Convention Center Follies.” U.S. cities invested tens of billions of dollars, expanding convention center capacity by 30 percent since 2000, while the demand for the space has barely budged.

All those consultants’ reports, it turns out, were based on optimistic assumptions and failed to anticipate the impact of industry consolidation and slower economic growth on the demand for meeting space. Even more curious was the consultants’ failure to take into account all the other cities contemplating subsidized expansions — something they surely knew because the same group of firms had prepared virtually all of the reports.

Rather than acknowledge their mistakes, however, the CIC convinced political leaders that the reason bookings had failed to meet expectations was that they didn’t have a big “headquarters hotel” to offer convention planners, who value such hotels because they reduce the cost and complexity of running such large events. Curiously, the private sector has been reluctant to seize on this golden opportunity to build them, so dozens of cities concluded that they had no choice but to provide subsidies for the hotels as well.

Once again, consultants cranked out studies projecting a big increase in bookings once the hotels were up and running. As Robert Nelson, a professor at the University of Delaware, notes in a recent paper, many of those studies also proved wildly optimistic. Some, like those in Trenton and St. Louis, wound up being sold for pennies on the dollar of debt. Others, like the city-owned and financed Hilton Baltimore, have lost millions each year since opening, requiring cities to dip into general funds to keep up with bond payments.

The August 2010 consultants’ report for Washington was refreshingly conservative in projecting the impact of a convention hotel. HVS Consulting and Valuation Services concluded that having a 1,175-room headquarters hotel would allow Washington to book 72,000 additional hotel room nights a year — the equivalent of three “citywide” conventions.

But it’s not clear that even that modest goal has been met. Although the city has been marketing the headquarters hotel since ground was broken at the end of 2010, the number of large “citywide” conventions booked over the next few years still hovers in the same 15-to-19 range as before the hotel was built. (Big conventions are typically booked five or more years in advance).

That’s not to say that having the hotel hasn’t allowed Washington to snare some conventions. Jim Goodman, vice president of the American Dental Association, said that without the Marriott Marquis, he would have to find rooms in dozens of distant suburban hotels to accommodate the 10,500 peak-night hotel rooms he needs to bring dentists to town. The ADA has committed to two Washington bookings.

Greg O’Dell, the director of the convention center authority, acknowledges that “it’s still too early to tell” whether the increase in convention bookings will be enough to justify the $206 million in public subsidy provided to the privately-owned Marriott Marquis: a $47 million grant from the convention center authority plus $159 million in forgone tax and lease payments. O’Dell blames the booking pace so far on a still-recovering economy and an “oversupply in the marketplace,” including competition from such places as Charlotte, Denver and Nashville.

Even if the convention hotel manages to generate those 72,000 additional room nights annually, they won’t have come cheaply — about $150 in subsidy for each room night, by my calculation. And that’s on top of the roughly $160 in subsidy the city was providing for the room nights the convention center was generating without the hotel.

And who pays those hefty subsidies? Well, there are at least two ways to think about that.

One is to look at who pays the taxes that support the bonds used to finance the facilities. In Washington, most of the money comes from the portion of the hotel, meals and rental car taxes that is diverted to the convention center authority. But convention-goers represent only 6 percent of the hotel visitors to the District who pay those taxes. That means the other 94 percent comes from tourists, business travelers and meeting-goers who never step foot in the convention center — as well as all of us Washingtonians who eat out and pay the higher meals tax.

Another way to think about convention subsidies is the way most economists would — in terms of the opportunity cost. This year, the center authority is spending about $95 million from dedicated tax revenue to subsidize convention business (operating losses at the convention center, marketing expenses, capital improvements and debt service). The city, however, could have used the proceeds from those taxes for other purposes, such as reducing income or property tax rates or giving more services to residents. Because of those forgone opportunities, it is Washington’s taxpayers who are really paying for the convention subsidies.

It is true, as convention boosters will argue, that $92 million in annual convention subsidies attracts several hundred million dollars of private spending by convention visitors, generating jobs for hotel and restaurant workers and taxi drivers, profits for hotel and restaurant owners, and tax dollars for the city.

But as Sanders documents, the convention industry routinely inflates those “economic impact” numbers, while the logic behind them is often flawed. In Washington, for example, most of those conventions are held during the peak spring and fall season when other visitors might well have come and spent their money instead.

Nor can it be said that the Marriott Marquis is just a convention hotel. If it is like other such “headquarters” hotels, it can expect to rely on convention-goers for only about a third of its sales. The rest will come from other business travelers, tourists and attendees at smaller meetings held in the hotel’s own meeting rooms. Some of that activity will be new business to Washington, but the rest will be lured away from other hotels in the region that are already losing bookings to their big, new, subsidized rival.

Although Marriott will operate the Marquis under a long-term contract, it doesn’t own the hotel. The owners, handpicked by Marriott, are Quadrangle Development and Capstone Development, a partnership whose big investor, in turn, is the government of Abu Dhabi. The developers and their investor contributed 60 percent — or about $330 million — toward the construction cost. And if the financial projections of HVS Consulting prove to be accurate, they will earn a respectable 14 percent annual return on that investment once finances are stabilized.

The real payoff, however, will come when they refinance the hotel in a few years, as is customary for such projects. In such a refinancing, the developers might get $230 million of their original $330 million back by replacing it with borrowed money. After paying to service that new debt, the hotel would generate enough cash flow to provide a 28 percent return on the on the $100 million of equity still invested in the property. That looks like a pretty sweet deal to me.

That’s also the way it looked to some city officials when they were negotiating with Quadrangle and Capstone back in 2010. That’s why they asked for a “clawback” provision that would have allowed the city to recoup some of its money if operating profits from the hotel exceeded a certain level or if the hotel was to be sold for a big profit. The issue, however, proved complicated and contentious, and the city’s higher priority was winning a guarantee that 80 percent of the hotel’s rooms would be made available for the big conventions. So the clawback provision didn’t make it into the deal.

And that illustrates the larger problem with convention center subsidies — that they tend to generate meager public returns and generous private ones.

So by all means, let us celebrate the beautiful new convention facilities we have built on Mount Vernon Square. At the same time, let us resolve not to dedicate any more public funds to a convention center arms race that no city can win.