ON THE MORNING OF MAY 22, 1986, Donald Trump, the New York real estate developer, called one of his executives, Anthony Gliedman, into his office. They discussed the inability of the City of New York, despite six years of effort and the expenditure of nearly $13 million, to rebuild the ice-skating rink in Central Park. On May 28 Trump offered to take over the rink reconstruction, promising to do the job in less than six months. A week later Mayor Edward Koch accepted the offer and shortly thereafter the city appropriated $3 million on the understanding that Trump would have to pay for any cost overruns out of his own pocket. On October 28, the renovation was complete, over a month ahead of schedule and about $750,000 under budget. Two weeks later, skaters were using it.1
For many readers it is obvious that private enterprise is more efficient than are public bureaucracies, and so they would file this story away as simply another illustration of what everyone already knows. But for other readers it is not so obvious what this story means; to them, business is greedy and unless watched like a hawk will fob off shoddy or overpriced goods on the American public, as when it sells the government $435 hammers and $3,000 coffeepots. Trump may have done a good job in this instance, but perhaps there is something about skating rinks or New York City government that gave him a comparative advantage; in any event, no larger lessons should be drawn from it.
Some lessons can be drawn, however, if one looks closely at the incentives and constraints facing Trump and the Department of Parks and Recreation. It becomes apparent that there is not one "bureaucracy problem" but several, and the solution to each in some degree is incompatible with the solution to every other.2 First there is the problem of accountability--getting agencies to serve agreed-upon goals. Second there is the