Revisiting the Argument that Government is Too Small

Once upon a time, academics argued over whether the government was too large or too small. I haven’t seen much of such conversations lately, even as government has blown far past its previous levels of magnitude and power. Perhaps that means it is worth reconsidering the major competing claims in the earlier discussion about the size of government, especially since recent changes have made the “government is too large” argument much more convincing.    

Arguments that government was too large centered around what Sam Peltzman described as “the incentives to use a political mechanism to redistribute wealth.” Arguments that government was too small relied on what John Kenneth Galbraith called the market sector’s “highly effective machinery for synthesizing private wants,” whereas “no similar process operates on behalf of the nonmerchantable services of the state.”

While the wealth-redistribution argument for the tendency toward over-large government appears valid, the opposing argument, correctly conceived, seems seriously flawed, and even more so now. A re-examination of the “government is too small” argument reveals that pervasive, deceptive “advertising” on behalf of government policies can accentuate the tendency of the state to grow too large, rather than constraining it to be too small.

The wealth-redistribution argument for governments to grow too large starts with voters who are rationally less informed about their political decisions than about their comparable private market decisions (termed rational voter ignorance). 

In making market decisions, people acquire information only as long as they expect the additional benefits from their making a better choice to exceed the additional costs of obtaining the information necessary to make it. People follow the same calculus in making their public-sector decisions, but the benefits are lower and the costs of acquiring the necessary information are higher than for individual market behavior. 

The primary benefit of becoming better informed is the extent to which it improves chosen results. This benefit can be substantial in market decisions, because decision-makers reap the gains. Because one person’s political vote is but one among very many, however, he has only a miniscule chance of influencing any outcome. For someone to investing in making a better-informed political vote yields virtually no personal benefits. 

The cost of acquiring the relevant information for public-sector decisions tends to be much higher than for individuals’ private-sector decisions. That is largely because political decisions require far more information than just the direct effects on oneself, and because much of it is of a type that requires scientific investigation to establish.

Contrast the knowledge required to make an informed decision about alternative national health insurance policies to that of selecting your own private health insurance plan. When selecting your own plan, you already know a great deal about your health history and at least a fair amount about your family’s health histories. Also, you need not worry about the effects of your choices on others, because your choices will not alter the market. But in considering national policies, you must consider all the health issues facing anyone, as they will have a sizable effect on the market. Further, you would need to know a large number of relevant elasticities of supply and demand, since that would affect the relevant magnitudes of the effects. And you would have to understand many other issues, as well as have some familiarity with the likely magnitudes, such as arise from adverse selection, moral hazard, technological changes in medicine, and so on. 

The combination of higher costs and lower benefits to becoming better informed means that most voters will know less about their political decisions, and will analyze what information they have less carefully, than they will for their market behavior.

Because the government can take income involuntarily from citizens (policies I object to can be imposed on me), in contrast to the mutually beneficial, voluntary exchanges in markets, rational voter ignorance motivates attempts by special interest groups to benefit themselves at the expense of the underinformed electorate. Groups whose members stand to gain substantially are rationally much better informed about those efforts and their effects than voters, leading to an over-large government more attuned to wealth redistribution than to advancing the “general welfare.”

Against the reasoning leading to an over-large government, perhaps the best-known argument in the other direction came from John Kenneth Galbraith’s The Affluent Society, cited above. He argued that “advertising operates exclusively…on behalf of privately produced goods and services. Since management of demand and emulative effects operate on behalf of private production, public services will have an inherent tendency to lag behind…demand which is expensively synthesized will inevitably have a much larger claim on income than…where no such influence operates.”

A strong case can be made, however, that Galbraith’s advertising bias claim, properly understood, complements the wealth-redistribution argument that implies an over-large government.

The argument that the advertising bias reinforces the redistribution bias toward an over-large government begins with Harold Demsetz’ recognition that “lobbying and public relations by government agencies and their suppliers and free press coverage of proposed government projects are at present functional equivalents of private advertising activity.”

In other words, the government actually engages in or controls tremendous amounts of advertising and public relations activity on its own behalf. Additionally, we must recognize that promoters of political “solutions” are far less constrained to tell the truth in their promotional claims than are private firms. So, as Demsetz saw, “It is not clear…that business firms have an undue advantage in want-creating activity. Washington enjoys free press and TV coverage and is immune from anti-fraud laws. Professional promoters are protected from the same laws by the camouflage of academic freedom.”

When politics’ far weaker “truth in advertising” constraint is combined with rational voter ignorance particularly the fact that many voters will evaluate the arguments for public sector actions much less carefully than they would private sector advertising — the result is that vast resources are devoted to misleading the public to oversell government “solutions” through plausible-seeming overstatements of the benefits and understatements of the costs of government programs. As Edgar and Jacqueline Browning summarized in their classic Public Finance and the Price System textbook, “Political speeches (as well as newspaper editorials) often rely on slogans, over-simplifications, inadequate theories, and misleading facts. Voters lack the necessary information to evaluate the assertions made by politicians, which in turn gives politicians little incentive to achieve accuracy and balance in their views.”

This enlarges the government domain, because the same rational voter ignorance that enables government “false advertising” makes it almost impossible to effectively combat its effects, since many voters won’t “waste” their time listening, reading, or thinking very hard about the true effects of government policies.

As Henry Hazlitt put it in the opening chapter (“The Lesson”) of Economics in One Lesson:

It is often complained that demagogues can be more plausible in putting forward economic nonsense from the platform than the honest men who try to show what is wrong with it.  

The reason is that the demagogues and bad economists are presenting half-truths. They are speaking only of the immediate effect of a proposed policy or its effect upon a single group. As far as they go they may often be right. In these cases the answer consists in showing that the proposed policy would also have longer and less desirable effects, or that it could benefit one group only at the expense of all other groups. The answer consists in supplementing and correcting the half-truth with the other half. 

But to consider all the chief effects of a proposed course on everybody often requires a long, complicated, and dull chain of reasoning. Most of the audience…soon becomes bored and inattentive. The bad economists rationalize this intellectual debility and laziness by assuring the audience that it need not even attempt to follow the reasoning or judge it on its merits.

Political salesmen advertise (in large part at taxpayer expense via such things as public hearings and constituent newsletters covered by franking privileges) every benefit for which even a semi-plausible sounding case can be made, tailored to an audience of voters who won’t think too hard about it, since they don’t anticipate any benefit from doing so. 

The result is that politicians and their supporters can tout policies with laundry lists of claimed benefits, because the cost of thinking up and distributing such a list is near zero. At the same time, they omit as many of the real costs as they think they can get away with. 

In contrast, the cost to opponents of constructing logically reasoned point-by-point refutations of such lists is astronomical, and the cost of getting rationally ignorant voters to pay attention is even higher. This is what ensures that a balanced rebuttal to false and misleading claims will tend to be ineffective at reaching voters, which in turn makes deceptive political advertising more successful, over-expanding government.

There have been a host of logical fallacies and errors committed in government’s ongoing misleading advertising blitz; so many that I will only briefly mention five of the most common here (but see this article for a bit more). 

Erroneous claims include that government spending creates jobs, when in fact it just moves jobs from ones individuals chose to those government dictates. Piling on to that misrepresentation is the claim that government spending creates multiplier effects, generating several dollars of total social benefits for each dollar spent, which assumes away the fact that the tax cost of financing such projects will have (always ignored) similarly multiplied adverse effects. Additionally, the same benefits are routinely counted in different forms, as if they were multiple different benefits, such as counting both jobs and incomes as benefits, even though jobs are actually the costs borne by workers in exchange for the income that is the benefit. In the same vein, what are really transfers from one party to another, generating no net benefit, are counted as if they were net benefits. In addition to these and other forms of overstating benefits from government projects, is the error of hugely understating the cost of a dollar of government funds as only a dollar, when the cost of the distortions introduced by the taxes necessary to raise the funds (which economists call welfare costs or excess burdens) add substantial additional costs to the burdens imposed. 

As even this very abbreviated list of misleading advertising by and for government indicates, the resulting tendency is to over-expand government. It works because the cost of self-promotional advertising to government salesmen is artificially low (given free by the media or paid by the taxpayers), because government salesmen are not effectively constrained to tell the truth, and because rationally ignorant voters evaluate their arguments less carefully than they would private market claims. 

Taken together, they form a highly persuasive case that, rather than any lack of advertising about the benefits of government programs leading to it being too small, the ability of the government to promote its solutions with large amounts of misleading advertising at low cost leads it to be too great to overcome. 

So why is the problem worse now than when people actually considered whether the government might be too large, rather than just doubling down on its expansion? Perhaps the best short answer I can provide is to cite an invitation I just received to an online Cato Policy Forum titled “Government Censorship by Proxy.” The core of the promotional blurb said:

During the pandemic, governments placed significant public and private pressure on social media companies to remove speech protected by the First Amendment, blurring the line between acceptable government speech and unconstitutional censorship by proxy. Concerns about this ‘jawboning’ only grew with the recent decisions in Missouri v. Biden finding that the pressure applied by various government actors likely violated the First Amendment…Join us as the panel discusses the options available to policymakers and why greater transparency is essential to combating such censorship.

In other words, the government has expanded its power, directly and indirectly, over what can and cannot be said on the dominant communication technologies of our time, and even stepped over the First Amendment to do so. That is about as far a cry as one could imagine from Galbraith’s “advertising operates exclusively …on behalf of privately produced goods and services,” so that “public services will have an inherent tendency to lag behind.”