On the LLP mailing list, Josh Kelly writes:
Does artificial government support of select industries (subsidization) lead to net economic gain on either a local, state, or federal level? I’m engaged with a friend over this issue, and while I don’t believe that any economic gain is incurred, he countered that with subsidization of an auto plant, the economy is stimulated in that other supporting corporations will soon set up shop in the region (rubber manufacturers etc), and thus it is of local benefit. I found myself unable to explain why subsidization is bad economic policy (outside of a philosophic view). Help?
Perry Eidelbus responds pretty effectively, but I thought I’d add a little bit of the sort of argument I’ve found helpful with these types of questions.
First, make sure you count both sides of the money flow. Let’s say the plant’s supporters want to give it a $100 million per year tax break for 10 years. That annual gift has to come from somewhere, and it will come from everyone else in the region in the form of taxes. (Pay no attention to the argument that it should come from government spending cuts. If cutting spending is a good idea, it’s a good idea with or without an auto plant.) Presumably the taxpayers had plans for that $100 million per year.
Some of them must be planning to start their own businesses, and those businesses will attract supporting corporations just like the auto plant. Add enough small businesses, and pretty soon you’ll get stores like OfficeMax, CompUSA, Kinko’s/Fedex, restaurants and coffee houses to serve the employees, convenience stores, gas stations, contractors to build and remodel the buildings, a produce market to supply the restaurants, cleaning services… Who can say that a few thousand small business owners couldn’t find better ways to spend the money than the auto plant executives?
(Thousands of little businesses aren’t as flashy and spectacular as an auto plant, and won’t get the mayor’s picture in the newspaper, but their dollars are every bit as good as the plant’s dollars and contribute just as much to the local economy.)
Second, don’t forget that there’s an opportunity cost to the auto plant. If it takes five years for a construction crew of 5000 to build the auto plant, you have to ask: What other things could those 5000 people have been doing for five years? They could also have been building schools and hospitals and homes and shopping malls. Why is it better to build an auto plant than all those other things? How do you know?
So where would the money would come from to pay those 5000 people if the auto plant isn’t providing it? Some of it would come from that $100 million that the taxpayers would have paid to the auto plant. The rest would probably come from loans and outside investors. After all, that’s where the auto plant was planning to get the money, and they won’t be needing it anymore. Ah, but what if they build somewhere else, they’ll need the money then, right? Yes, but the people who would have been building all the schools and hospitals and homes and shopping malls in that location will no longer be needing their loans and investment money, so it will be available. This why we have capital markets.
Third, people respond to incentives. If you bring a factory with 30,000 jobs to your town, everybody knows it. People will move to your town just to get jobs. They’ll need housing and grocery stores and funeral homes and everything else people need. There will be a boom. Your town’s economy will grow. But that doesn’t mean that the economy of any individual family will grow. There may be more wealth, but there are also more people sharing it. It’s like when companies merge: The company is bigger, but the workers’ paychecks don’t get bigger.
And so on. The idea is to make the scenario specific enough that you can keep track of all the flows of money, material, and labor and account for both sides of every transaction. If they want to build a factory, ask where the materials, labor, and land came from, and what could have been done with them instead. Or ask where the money came from and what it cold be used for instead. (For that matter, if they want to build a park, ask what happens to the 300 families who could have built homes there.) If you prevent your opponent from smuggling in benefits without counting the cost, most of the claimed benefits of subsidies will go away. Also watch out (in your argument and your opponent’s) for double counting. You can count the money, or you can count the things the money buys, but you can’t count both. (I’ve been a little fast and loose with that here; my three major points overlap.)
I guess that when you keep track of this stuff and explore the details, it all gets pretty complicated. It becomes hard to tell which course of action will produce the best result. It’s hard to pick the winners. And that’s the point: There’s no reason to believe that government officials can pick winners any better than the free market can.
(Sorry this was so long. I didn’t have time to make it shorter.)