Exactly 40 years ago we heard a computer voice intone: “Strange game. The only winning move is not to play.”
Okay, spoiler alert: that’s the climactic moment of the 1983 movie “War Games,” in which an AI system “learns” the concept of futility, and then “plays out” all the conceivable strategies of triggering mutually assured destruction in nuclear war. The only winning move is to avoid playing in a strategic setting that conforms to the prisoner’s dilemma in the first place.
Public Choice economists in the 1970s came to an important recognition: While competition is good, and can be very good, in some institutional settings—commercial society, international trade with division of labor in market systems generally—in other settings competition can be harmful. In fact, under some circumstances competition can be devastating, actively using up or destroying far more resources than are collected by the putative “winner” at the end of the game.
The problem is recognizing the kind of institutional setting where we find ourselves before trying to devise a strategy that will allow us to “win.” The insight of Public Choice is that it is the devising of the “rules of the game,” rather than the playing of a particular game, that lies at the heart of good public policy. In their underappreciated book—if a book that has been cited nearly 3,000 times can be called “underappreciated”—Reason of Rules, Geoffrey Brennan and James Buchanan lay out a logic of constitutional choice that can help ensure that societies avoid systems in which the only sensible move is not to compete (“play”) in the first place.
It is necessary to separate the process through which the rules are determined from the process through which particular actions within those rules are chosen. Again, however, the distinction is somewhat more difficult to draw in the social setting because of the complex interdependencies between the rules that define the constraints on private behavior and the rules that define the constraints on the political agents who may engage in activities involving changes in the first set of rules. That is to say, legislative majorities may be acting within the rules (the political constitution) that constrain their own behavior in changing the rules that constrain the behavior of persons in their private capacities. One must be careful to make the distinction between a choice among rules and a choice among strategies within rules applicable to the situation confronted by a well-defined decision-making unit.
Legislatures have two jobs, at least. One is to devise the rules under which competition takes place. The other is to win that competition, as individual members who seek reelection and as parties that try to command majorities and vote in policies they favor. The temptation, typical of “prisoner’s dilemma” settings, to win a particular round of the game may overwhelm the clear knowledge that for many “strange games” it would better for everyone not to play at all.
Strange Game: State Development Incentives
One such strange game is the economic development bidding wars routinely fought by state and municipal governments. No one steps back and asks the obvious questions: Should state governments use economic incentives to “compete” for new investment and construction of manufacturing plants in the first place? Do bidding wars for “economic development” deliver the benefits that politicians claim for those programs? And how would we know?
The temptation is simply to play the game as it is given. If incentives—which would be bribes or kickbacks if undertaken by any private entity—are offered as a way of affecting the siting choices for a new Mercedes assembly plant, or a new chip manufacturing facility, what is the net effect? Imagine that there are two states, Mungeria and Salsmania, with essentially identical tax policies, work force quality, and infrastructure. The state governors are approached by QAI-Marsh, a company that manufactures quantum AI-powered marshmallow candy bars, the hot new thing among influencers world-wide.
Now, QAI-marsh pays wages that are 25 percent over the average state wage, and it expects to create at least 12,000 new jobs, and build a large factory which will require hundreds of construction jobs that would last for several years. If a state can attract a QAI-marsh plant, it will likely also attract other businesses, and the governor and majority party in the state legislature can likely win the next election in a cake-walk.
The problem for Mungeria is that Salsmania is going to bid, also. Suppose that the benefits to the state, including the usual hand-waving about “multipliers” and “good jobs” are not just smoke and mirrors (they may be smoke and mirrors, but suppose not). How much will the two states “offer,” in terms of tax breaks, direct cash grants, and subsidies to build infrastructure such as roads, electrical capacity, and so on?
The optimistic conjecture is that the “winner” will bid the full amount of the estimated benefit to the state. Suppose that adding up all the multiplier, construction, and other increased economic activity in the state implies a “benefit” of $320 million. Mungeria and Salsmania will each offer packages—differing in detail, but not in value—of something close to $320 million.
That’s the best-case scenario, folks. And even then literally ALL of the economic benefit that would have gone to the state is given back to the corporation as a kickback. Of course, there are three things (at least) that can go wrong:
1. The promises of QAI-marsh turn out to be unreliable, and the half-built facility is abandoned, having wasted tens of millions of dollars of taxpayer money, but costing QAI-marsh very little. Further, the “incentives” are really tax money taken from firms already in the state, for the right reasons, and transferred to outside firms in a kind of political prostitution.
2. The “benefit estimates” turn out to be wildly optimistic: Even though QAI-marsh does build and operate the facility the actual benefits are only $200 million. State taxpayers have wasted $120 million because some glib consultant had learned to say “multiplier” in forecasts that had no actual economic basis.
3. Worst of all, there is no reason to expect that the upper limit on bids will be constrained by the actual economic value (and see #2, above, because “actual” economic value is probably a fib in the first place!). But suppose the benefit to the state really would be $320 million. It’s not the governor’s money, and it’s not the taxpayer’s money. Remember, the goal of attracting the facility is political, not economic; if Mungeria bids $450 million, then QAI-marsh says “Yes!”, builds the facility, and the political goal of electoral success is secured. Far from being a harm, overbidding is actually a benefit for political leaders doling out other peoples’ money.
What all this means is that it is better, from the perspective of state taxpayers, “not to play at all.” But how could this be accomplished, since the game is one that politicians love to play, even though it harms the interests of the citizens they have pledged to serve?
There are three possibilities, it seems. The first is a statute that prevents the payment of grants and tax incentives as part of a recruitment strategy. But as Brennan and Buchanan pointed out in the quote earlier in this paper, that is simultaneously a change in the rules of the game—“no incentives, don’t play at all”—and a play of the game—appeal to voters to win the next election. It appears that voters like, or at least reward, politicians who do play the incentives game, so that’s a fail.
The second is a restriction on the definition of “public purpose” or value for citizens. It could go along an equal protection route, requiring that states cannot discriminate on tax policy, but must treat all corporations the same way. Or it could be some other general requirement that prevents specific, targeted private benefits using public funds. My home state of North Carolina actually had such a set of precedents, and the rules seemed to be holding, until the catastrophic state supreme court case in Maready vs. City of Winston Salem 1996. That decision redefined “public purpose” away from objective, widely shared goals and moved toward what has since come to be called the “rational basis” rule, meaning that almost anything that the state legislature wants to do, it can do, because the legislature decides what’s good for the public. The problem, as Brennan and Buchanan argued nearly 40 years ago, is that the incentives of politicians and the goals of citizens can diverge, and nowhere is that divergence more obvious than in the case of targeted, private benefits to specific corporations paid for at public expense.
The third possibility is constitutional restrictions, with a state constitutional provision outlawing the excessive use of economic incentives for political purposes. While there are examples of such restrictions, they are increasingly seen as archaic among policy analysts and unpopular for political elites in state capitols who are looking to improve their ranking as a “business-friendly state.”
I don’t think this is hopeless, but it’s hard to be hopeful in the current environment of corporate cronyism and state boosterism. The idea that the legislature is fundamentally responsible to reshape the rules to make systems work better, rather than to manipulate the current system for short-term political advantage and electoral victories, seems out of fashion. Nonetheless, for citizens, the only winning move is not to play at all, and not to reward the politicians who do.