by Filip Palda
Public choice scholars, especially those adhering to a branch of the field called constitutional political economy, believe that it is during the first stage of the political game that scholars can make their influence felt. They are ready to concede that during the second stage, the chance for scholarly influence may be small because at that stage the pressures of constant competition render the individual irrelevant.
To understand this striking way of reasoning let us then examine public choice to see what its ideas about competition and political efficiency are.
Public Choice
PUBLIC CHOICE IS based on the median voter theorem. The theorem holds that in political equilibrium parties will follow the wishes of the “median” or middle voter.
Take a room with a hundred and one people and ask them what their desired level of government spending is. Arrange them in a line starting at one end with those wishing the lowest level of spending to those wishing the highest at the other end. The fifty first person in this identity parade has fifty people to her left and fifty to her right. She is at the “center of mass” or median of the distribution of people arranged by their preferences. Her preferences are those around which parties will shape their platforms.
In 1929 Harold Hotelling solidified this notion in his article on where firms would locate to be closest to customers. It makes no sense for each owner of two ice cream stands to locate at far ends of the beach. If one locates at the left end the other can nab all customers by coming between her and everyone to the right of her. To keep customers, she will move her stand just to the right of her competitor. Both will keep leapfrogging each other until half the potential customers are on one side of each stand. Similarly, parties “located” on a left-right political spectrum will avoid extremes and apply the logic of ice cream stands. Move to the center to nab the most voters.
The rush to the center produces a political equilibrium based on the optimizing calculations of parties and of voters. In 1948 Duncan Black reproduced Hotelling’s result without citing him. In 1957 Anthony Downs reproduced the Hotelling result, citing Hotelling but ignoring Black. This sort of selective memory in citation is normal business in academia, perhaps because few people care about intellectual history and just want to get to the result. The result is now widely known as the median voter theorem.
The median voter theorem is more specific than Becker’s equilibrium model of politics because it posits a spectrum along which policies are formulated, and makes some detailed assumptions about voter preferences such as “single-peakedness” which need not concern us.
This more focused picture of politics in no way contradicts Becker’s model. Instead, the median voter model can be seen as a subset, or one possible manifestation, of equilibrium that fits into Becker’s all-encompassing formulation of that concept. The fact that one model snuggles inside another alerts us to the possibility that long before the Chicago school, notions of efficient political equilibrium were stirring.
Stirrings of political equilibrium and efficiency
IN THE 1940’S and 50’s when the concept of political equilibrium was just beginning to emerge, little thought was given to its efficiency. The best brains in economics at the time, such as Lionel McKenzie, Kenneth Arrow, and Gerard Debreu were struggling with proving the efficiency of economic equilibrium. Düppe and Weintraub describe this quest in their 2014 book Finding Equilibrium. Political equilibrium and its implications for efficiency were still undeveloped concepts. Yet something political stirred at the fringe of economics.
In 1954 Paul Samuelson ventured into the then nascent field of public finance economics. He asked whether the rules of free markets could apply to governments. The answer was no. The sole economic justification for government at that time was that it should provide “public goods”. Private markets would not produce enough of such goods because they were “non-excludable”. You could not prevent someone from consuming them.
Streetlights are an example. How do you charge someone for enjoying the safety of walking down a lit street? The non-excludability of lighting explains why most cities went unlit until the mid-nineteenth century.
Private suppliers would not produce lighting because they could not charge for it. Charitable suppliers could not produce lighting because of the problem of free riding. Some good souls might contribute to a street-lighting fund, but others would simply not pay and coast on the efforts of others. Such ventures generally founder, despite an important truth. It is to the benefit of all, except perhaps cat burglars, that streets be lit. Yet without the means to enforce payments, a project that benefits all goes unrealized. This lack of coordination between members of society creates a Pareto inefficiency. Enterprises of great public value never see light.
To produce such “non-excludable” goods society needs a system of forced payment. It is called taxation. Such a system must itself be efficient. If government takings destroy the economy, there is no value in producing public goods. It was only around the mid-nineteenth century that western governments figured out how to extract resources from their subjects efficiently. This efficiency allowed governments to ponder the other side of the decision ledger, which is, once you decide a non-excludable good can be financed at some reasonable cost, what level of such a good should you provide?
Economists have a ready answer to such questions. Benefits net of costs are maximized when the marginal benefits and costs equal each other. In economics, as in mathematics, marginal means extra. A project should be carried to the point where the added benefit of spending an extra dollar on it just equals the added costs. If the added benefit is greater then more should be spent. If it is lesser, then less should be spent. This decision rule can only be of practical use if governments know what a project costs and the benefits it brings.
Knowing cost is simple, but not as simple as you might imagine. There is naturally the direct dollar cost to consider, but each dollar of taxation discourages economic activity. The value of discouraged activity is known as deadweight loss.
The benefits of a project are generally harder to divine. Governments seeking to maximize Pareto efficiency must push production to the point where the extra amount that the sum of people in society is willing to pay, equals the cost.
We do not need to get into peoples’ heads to scan how much pleasure they get from something. Pareto efficiency is concerned strictly with unexploited opportunities. People demonstrate their willingness to pursue these opportunities through their willingness to pull money out of their pockets.
But how much money will people pull out for non-excludable goods? This is a hard question because there are no markets for non-excludable goods. Thus we have no idea of how much extra money people would be willing to pay for more of the product. The answer lies in an egg-hunt for market data of similar goods but which are excludable, and need not concern us here lest we embark on a distracting detour from the story of political equilibrium and efficiency. The point is that there are imperfect means of calculating the marginal benefits and costs of providing non-excludable goods. But knowing how to calculate these costs and benefits did not satisfy economists that they could devise a rule for optimal government intervention where markets failed.
The rule economists finally settled upon arose from a mysterious obsession they nurtured in the 1950s. For some reason, economists came to think of public goods of as possessing a second feature in addition to non-excludability. The infelicitous term for this second feature was “non-rivalry” in consumption. Public lighting is non-rivalrous because my enjoyment of this lighting does not detract from anyone else’s enjoyment. A sandwich is rivalrous because the bite I take is usually no longer available to others.
Non-rivalry was not a necessary justification for government intervention. Movie theatres provide a non-rivalrous product but manage without government aid because they are able to keep away non-paying viewers. Only non-excludability is a justification for government’s intervention to push an economy to the
frontier of Pareto efficiency. Perhaps economists of the time believed that non-rivalry and non-excludability were inseparable.
In his essay on light houses Ronald Coase gives some idea of how people were thinking at the time. Let us not quibble. Our quest is to understand how Paul Samuelson’s result on the optimal provision of public goods ignited the fusion of thought on political efficiency and equilibrium.
Samuelson’s result
SAMUELSON SOLVED THE problem of the optimal provision of public goods by recognizing that a non-divisive good brings some benefit to each person each time the quantity of the good is increased. This extra or “marginal” benefit may vary from person to person. To get the optimal level of public good one simply had to keep increasing the level of good until the money and deadweight loss cost of the tax to finance this marginal quantity equalled the sum of marginal benefits to all who received non-rivalrous benefits.
The result interested economists in the field of cost-benefit analysis, but it had a broader significance than that imagined by social engineers seeking a mathematical formula for optimal government intervention. Samuelson’s result could be wedded to the fundamental equilibrium concept of public choice called the median voter theorem. The child of this wedding would be the first ever formal proof that politics could be Pareto efficient.
The proof of this assertion is disarmingly simple. The median voter theorem posits that parties will offer a platform appealing to the median voter. The median divides people on the left equally with people on the right. If public opinion is distributed evenly from left to right on some index of issues, then the median is also the average of opinions. But what is an average other than a sum, divided by the number of individuals being counted? Samuelson’s theorem says that the optimal level of public goods is determined by the sum of marginal benefits received by the populace made equal to the marginal tax cost. One can divide both sides of this equation by a constant, such as the number of people and still obtain the same optimal level of public goods. The average marginal benefit however is also the benefit coming to the median voter. Thus on one side of the equation one can forget the preferences of all voters except that of the median voter. Hence the median voter’s preferences, balanced against the tax cost of providing the public good determines the optimal level of government intervention in the economy.
The efficiency of government in a society where the median voter rules emerged decades ahead of Becker’s fusion of economics and politics. The notion of public goods was still too fresh to be widely known. The concept of political equilibrium was barely fledged. The academic world was not ready to appreciate the Black-Samuelson synthesis in which political efficiency of the rule of the median voter were fused.
Not until the mid-1970’s did economists start pondering this result. They came from the public choice school, which at that time held government to be inherently inefficient. Charles Rowley explained: “If problems of monopoly, externalities, public goods and bounded rationality afflicted private markets, they simply ravaged political markets that confronted individuals with massive indivisibilities and severely limited exit options.”
They need not have worried too much. As Dennis Mueller explained in this 2003 survey of public choice “the sophistication and elegance of the theoretical models of public choice far exceed the limits placed by the data on the empirical models that can be estimated”. What this means is that to get empirical support for your theory you need data that vary sufficiently in order to reveal relationships between dependant and causal variables. Efforts to validate the median voter model focused on trying to show that government spending in a political district is more closely wedded to the income of the median voter than it is to the income of the average voter.
The data do not show enough variation to allow the dominance of median income as an explanatory variable to be discerned. The median voter may determine the level of government spending, but conclusive evidence is still being vigorously debated. Most importantly, we cannot conclude that politics leads to an efficient provision of public goods in the narrow Black-Samuelson sense because we must still proceed with caution when arguing that equilibrium is determined by the median voter.
A summary of differences
OUR TOUR D’HORIZON of public choice explains why it opposes the notion that political competition leads to government efficiency. The data do not give any consistent support for this notion. This is what one would expect if one accepts the public choice view that competition is difficult in politics because of voter ignorance and powerful barriers to competition that power-seekers put in place.
The Chicago riposte to this critique is that competition must be viewed as taking place within unalterable constraints. Chicago admits that voters choose to be selectively ignorant because it does not pay to be fully informed. Only one party may rule at any given time and is thus shielded from the simultaneous offering of competing government products by political rivals. Other facts such as the large fixed costs of attaining power create barriers to entry into politics. But there is no point in whining about these realities. They are constraints.
Everyone maximizes their wealth within constraints and that is the best we can do. Mainstream economists may say that the minimum wage is an excrescence from politics that creates unemployment. Chicago would say that, yes, this may be so, but if the minimum wage exists to transfer income to the poor, it does so because other means of transferring income, such as through taxation and direct subsidy, create greater deadweight loss. To push the point to its extreme, Chicago would give a nuanced answer if asked to judge the Meso-American practice of human sacrifice. To us such behaviour is shocking. But it may have been a method to control population growth that avoided the deadlier alternative of civil war due to excessive competition for the control of nutritive resources.
No one summarizes the Chicago view better than 18th century poet Alexander Pope. In his Essay on Man he wrote “All Nature is but Art, unknown to thee; All Chance, Direction which thou can’st not see; All Discord, Harmony not understood; All partial Evil, universal Good: And spite of Pride, in erring Reason’s spite, One truth is clear; “Whatever Is, is RIGHT.”
The Chicago argument for political efficiency maddens public choice scholars. They suspect themselves trapped in the vise of a sophisticated tautology: in politics only the most efficient outcomes will obtain because the human effort to leave no opportunities for the creation of wealth unexploited leads to efficiency. The tautology lurks in the phrase “the human effort to leave no opportunities for the creation of wealth unexploited”. This can be logically equated to meaning that “human effort leads to efficiency”. Thus the Chicago proposition could be restated as “in politics only the most efficient outcomes will obtain because the fact that human effort leads to efficiency leads to efficiency.” The evolution of species by natural selection of the fittest and Freudianism provoke similar irritation in their critics.
Despite their ire at the Chicago view, public choice scholars quickly learned to play a similar game. In his 1982 essay on the Chicago school, Melvin Reder called this game “tight prior”. You build up a theory on the basis of ideology. There is no such a thing as letting the facts speak for themselves. There are too many facts to make sense of them by sifting them. You have to assert your view of the world and see if the facts go along (some people call this putting on rosy glasses). Chicago believes that equilibrium prevails in economic and political markets and that this equilibrium is Pareto efficient. Contrary evidence is not particularly welcome. As Reder writes “Any apparent inconsistency of empirical findings with implications of the theory, or report of behavior not implied by the theory, is interpreted as anomalous and requiring one of the following actions: (i) re-examination of the data to reverse the anomalous finding; (ii) redefinition and/or augmentation of the variables in the model.” Along this line, government efficiency is difficult to measure as is political competition. Until such reliable measures come along it will be diffi
cult to reject the tight prior view that political competition leads to government efficiency.
Given the challenges of finding conclusive evidence on the link between competition and efficiency, public choice scholars are free to put forth their own tight prior. Yes, competition in politics may lead to economic waste from “rent-seeking”; the efforts expended on dividing wealth rather than on creating it. Yet this is not an unalterable optimum as Chicago maintains. Public choice scholars can give advice to move politics towards greater efficiency. These scholars do not sit outside politics but rather are an organic part of it. They play a decisive role in determining the course of society.
In the vacuum of decisive empirical evidence, the argument between Chicago and public choice appears to be a metaphysical debate going back to the time of Plato. Do we have free will? Chicago seems is obdurate in its “no”. Public choice says yes, but in a subtle manner that poses a significant challenge to Chicago thinking.
Enter Buchanan and Brennan
IN THEIR 1977 article Towards a Tax Constitution for Leviathan Nobelist James Buchanan and Geoffrey Brennan argued that conventional economic thinking about the rules that should govern optimal government intervention were blinkered. Conventional thinking imagined either a benevolent social planner, or a political system geared towards the wishes of the median voter. In such a system governments would levy taxes that minimally discouraged the creation of wealth and would spend only to the degree that public goods were provided optimally for the populace.
Buchanan and Brennan rejected this idyllic view. They feared that government would ignore economic notions of optimality and instead suck as much tax as it could from its subjects. The growth of a government “Leviathan”, a mythical beast rising from the ocean to devour ships, would then lead to all sorts of abuses of power. To contain Leviathan, they argued that politics should be seen, in the fashionable jargon of the time, as a “two-stage game”.