A number of readers have asked what I think about New York Governor Andrew Cuomo's call to sell the Javits Center in Manhattan and build the world's largest convention center out in Queens. If this is a good idea—which it probably is—then it's simply because the merits of selling the Javits Center exceed the demerits of building a new convention center.
I may expand on this for a longer piece at some point, but the question about convention centers is where's the market failure? If the benefits of additional demand for hotel rooms exceeds the cost of building the convention center, then why don't hotel operators just build the convention center? Indeed, you see that it's extremely common for hotel operators to integrate conference facilities into their proprerties for exactly this reason. These facilities tend to be relatively small compared to convention centers because, presumably, that's what makes the cost and revenue curves line up. Someone needs to explain not just why convention centers have ancillary benefits, but why those benefits are so unusually hard to internalize that they justify subsidies. I would say that in high-rent cities, it's actually residential housing that has the hard to internalize benefits since actual residents are a good deal more mobile than conventioneers. Meanwhile, many cities—including the one I live in—appear to me to have a perverse policy of taxing hotels heavily and then deciding that the tax revenue associated with hotel occupancy justifies subsidizing hotel construction.