You know, it’s easy to understand why officials from Caldwell County in western North Carolina wanted to spend millions on incentives to attract a Google computing facility. Caldwell is small and poor, and most of its talented young people leave to go to college and don’t come back. Still, you have to think about whether it’s worth it to spend over $1 million dollars per job in incentives to attract a building that’s going to employ about 200 people. If you truly believe it’s going to attract additional growth, you have to think about what the mechanism is going to be. If there isn’t a reasonable prospect that the development is going to take, you’re better off hiring people to dig holes all day at $100,000 per person (or better still, offer $100,000 grants to researchers).

The mayor of Lenoir points to the BMW plant in Spartanburg, South Carolina as an example of how the Google plant might work, but he must see the difference. As the N&O piece linked above points out, there are no manufacturing linkages involved in the case of the Google center. There is no reason to expect that the center will attract a larger pool of talented labor that might be used to draw additional companies.

There are two lessons incentive fans must learn. The first is that attracting manufacturing is great for the short term, but unless you’re producing innovations that will sustain growth or a pool of knowledge workers than can operate outside a limited manufacturing base, that development is not going to stick. The Greenville-Spartanburg area is doing well now, but it lacks flexibility and will find itself in dire straits if BMW scales down operations.

The second is that in ANY development situation, one firm won’t do the trick. The reason the Research Triangle took off is because it could attract companies based on an existing pool of knowledge workers (thanks to the local universities) and because it sought to draw many different firms. Once you create the labor pool, the process becomes sustainable–new companies come for the human capital and new human capital comes for the companies. But one won’t do it. You bring in the workers you need and the music stops. Worst of all, you have to overpay in incentive terms, because companies that rely on and produce spillovers would rather be near other companies. To get them away from RTP or Silicon Valley, you have to compensate them for the loss of not initially being around their peers.

It sucks that not every place in the country can be economically vibrant, but that’s the nature of things–businesses like to concentrate. If, after recognizing that, you still want to pay $1 million per job, then that’s your prerogative as a government. But you’d better make sure tax payers know the deal.