In a recent New Yorker article which also attracted attention from APM’s Marketplace program, Jill Lepore wrote that taxes have never gotten good press. The best defense for taxes she could find was one that’s almost a century old, by Supreme Court Justice Oliver Wendell Holmes who wrote that “Taxes are the price we pay for a civilized society.”
Those who’ve never taken an economics course might associate economics with conservative ideology and think that economists wouldn’t care much for taxes. But economics actually provides a perfectly clear justification for them. Economists consider taxes a necessity.
Consider that some goods or services are best provided to a large population without ready possibility for discriminating among those helping and those not helping to cover their costs. A city’s streets provide an example. At least until the arrival of modern computer chips, there was no way to tax individuals based on their specific usage of particular streets without putting toll collectors on every corner and slowing traffic to a crawl. Or think of enforcement of safety regulations protecting our food supply. Companies could voluntarily adopt safety standards and tell customers that they were doing so, but how can individual customers verify that the standards are really maintained? So there’s a collective demand for safety enforcement on the part of most consumers, but no practical way to get each individual consumer to pay a share calibrated to her consumption of specific products.
Collective provision of road repair, safety inspection, street lighting, combating crime, limiting emission of pollutants, and many other services, is efficient, but not all individuals can be counted on to assess themselves an appropriate share of the financial responsibility and to voluntarily pay up. Taxes are simply bills with which each of us are presented, payment of which helps cover the costs of our public services. They differ from the requests for contributions that we receive from charitable organizations because rather than being voluntary, our contributions in the form of taxes are legally mandatory. Assessing how much it’s appropriate for each to pay is left to be addressed by the political process.
What do taxes have to do with “civilization?" The simplest answer is found by thinking about law and order. While we achieve some order by socializing our children to take aboard some basic norms of behavior, no modern society can function without police, prisons, and courts. These too are public goods financed by taxes. It’s thanks to those taxes that we have relative security of property and of person, as well as the things that go under the heading “rule of law.” It’s this package of items that I imagine Holmes had in mind when using the term “civilization.”
The shift from voluntary to mandatory contributions to cover the cost of public goods is the most straightforward part of the tax equation. Economic theory holds—more or less correctly—that if paying in to the fund for providing public services were strictly voluntary, most people would find reasons to avoid paying their full shares. After all, almost anyone can think of other things to do with the money, and if paying taxes were strictly voluntary, most would probably prioritize those other things. By giving government the power to impose penalties on any who fail to pay, we cut through the potential “free rider problem” of non-payment.
In reality, though, there continues to be a voluntary element in tax payments. Many individuals and companies can under-declare their earnings and over-declare their tax deductions, and the government lacks the resources to investigate each one with care. Random inspections coupled with truly draconian penalties—e.g. if you knowingly underpay, you’ll be shot—could solve the problem, but in civilized countries we rule out penalties large enough to fully deter tax avoidance out of self-interest. We count, in fact, on the voluntary willingness to pay of most of our population. The stated tax rate serves as an anchor stipulating the amount of obligation, and the potential penalties serve as a partial deterrent, but desire to avoid the stigma of being caught cheating and our internalizing of the idea that paying our share is the right thing to do also play large parts in explaining payments. The evidence includes calculations showing that the expected monetary cost of tax avoidance (probability of being caught times size of penalties) is usually insufficient to induce payment by itself. Also, studies show that levels of tax compliance differ widely across countries. These differences are better explained by differences in public attitudes towards the democratic process and the legitimacy of government institutions, as well as beliefs about how many other people are compliant, than by differences in likelihood of detection and in size of penalties.
So justifying taxes in general—the assessment of payments for public services, and their formal designation as mandatory—is easy. The hard part is agreeing on what and how much government should do and how much given classes of individuals and companies should pay. Here, there are legitimate reasons for debate, and no one answer trumps others on pure efficiency grounds. The best that can be said is that by embracing the principle of democratic choice, we’re hoping that we can at least minimize the sum of unhappiness from the decisions adopted. For each person who feels that the public is being asked to pay for too many government services, that is, there will be another who feels that the services provided are too few. Positions may also vary on an issue-by-issue basis, for example I may feel that we spend too much on tax breaks for oil companies and too little on public health measures, and you may have the opposite view. A democratic political process tends towards the median outlook along each spectrum.
The biggest area of controversy over taxes in the U.S. in recent years has concerned the use of taxes and public assistance programs to effectively alter the distribution of income and of economic well-being. Once the idea that those with the highest incomes can afford to pay a larger share is out of the box, the tax system can begin to play a redistributive role in addition to its original role of solving the free rider problem of funding public services. It turns out that income distributions in all market economies are characterized by having most incomes clustered around a fairly low mode with smaller and smaller numbers of individuals earning high, very high, and even higher incomes. A consequence is that if one were to propose that those earning above the numerically average income pay a low percentage of their income as taxes while those earning above that average income pay a higher percentage, there wouldn’t be 50% benefiting and 50% losing out from the arrangement, but actually a much larger fraction of the population benefiting than losing. This is because the numerically average earner is earning far more than the median earner—the person who half earn more than, half less than. And this is the starting point for understanding why combining democracy with a tax system the details of which are to be determined by the political process is so likely to lead to some pressures for tax progressivity, which partially turns taxation and public expenditure into tools for altering the patterns of inequality produced by the forces of the market.
In sum, the desirability of taxes is not at all in doubt, at least among economists. But who should pay how much is a thornier matter that must be hashed out in the political process. I’ll return to the debate on tax progressivity in my next post.