Why Taxing Consumption?

© Springer International Publishing Switzerland 2015
Helmut P. Gaisbauer, Gottfried Schweiger and Clemens Sedmak (eds.)Philosophical Explorations of Justice and TaxationIus Gentium: Comparative Perspectives on Law and Justice4010.1007/978-3-319-13458-1_7

7. Why Taxing Consumption?

Justifications, Objections and Social Cooperation

Xavier Landes 

Department of Media, Cognition and Communication, University of Copenhagen, Karen Blixens Vej 4, 2300 Copenhagen, Denmark



Xavier Landes


Robert Frank is famous for proposing an incremental tax on consumption. His proposition is motivated by the control of positional externalities, i.e. the costs that individuals impose on each other when they consume goods for securing or acquiring social status. A close analysis of Frank’s proposition identifies three justifications for a tax on consumption: efficiency, paternalism and equality. This chapter has two purposes. Firstly, it reviews these justifications, highlighting some objections and possible replies. As such, it suggests that reasons based on equality or paternalism are controversial while the invocation of efficiency is actually grounded in an underlying view of social cooperation. Secondly, this chapter advances the idea that an ultimate justification for the choice of specific tax base (consumption, income and wealth) expresses such an underlying view. In other words, the choice of a specific tax base is not totally instrumental, it has some intrinsic moral value too. In this respect, the chapter ends with a comparison between taxing income and taxing consumption. It is shown that a tax on consumption raises questions that should be answered by political philosophers.

I am thankful to Bruno Verbeek and Daniel Halliday for their comments and inputs as well as to the participants to the conference on Justice, Taxation and Social Policy organized by the Center for Ethics and Poverty Research at University of Salzburg.

7.1 Introduction

Taxation is the price which we pay for civilization, for our social, civil and political institutions, for the security of life and property, and without which, we must resort to the law of force.

Journal of the House of Representatives of the State of Vermont, October Session, 1851

Taxation is a key component of public policy by generating resources for public goods, social insurance and redistribution. Decisions about taxation may also affect political stability by nurturing social unrest and revolutions. Examples are abundant: the American Revolutionary War started with a tax controversy (the 1765 “Stamp Act”), like the French Revolution.1 In United Kingdom, the introduction of a poll tax (“Community Charge”) contributed to Thatcher’s fall in 1990.

Despite the importance of the topic, it has attracted little interest in political philosophy. This is striking when compared to the refined analyses in distributive justice about the use of resources extracted from taxation. Various reasons can explain such neglect. One might argue that no matter how resources are collected, the only relevant question is what to do with them. There is also the view shared by many egalitarians that just contributions are contributions that mirror one’s ability to pay or produce (e.g. talents in Rawls), which is supposed to end the discussion.

However, unless one controversially assumes that how public resources are collected does not matter at all, there is space for investigating the normative dimensions of taxation. It is difficult to consider that, for instance, the choice of taxing consumption instead of wealth or income, or to impose a progressive rate instead of a flat one is normatively indifferent. If this point is accepted, a normative theory of taxation has to articulate the following dimensions.


Why taxing? What reasons do public institutions have for collecting part of, what looks like at first sight, individuals’ resources?



What taxing? What is the “right” tax base, consumption, income or wealth?



How taxing? Which rules should be followed when implementing citizens’ tax duties?


I address here the second question through a discussion of Robert Frank’s advocacy for a progressive tax on consumption (Frank 1999, 2007, 2008, 2011).2 The methodology is to present Frank’s position (first section) and to evaluate three reasons that might support a consumption tax: efficiency (second section), paternalism (third section) and equality (fourth section). When discussing these reasons, income or wealth taxes are not mentioned for the sake of a comparative analysis, but for highlighting the strengths and weaknesses of Frank’s position. The contribution ends with a brief exposition of the core of the issue: the conception of social cooperation embodied in the choice of a tax base (fifth section). In the fifth section and the conclusion, I claim that political philosophers, especially those interested in distributive justice, need to be more concerned with taxation.

7.2 Positional Consumption and Externalities

Inspired by Thorstein Veblen’s seminal work on conspicuous consumption (Veblen 1994), Frank advocates for a steeply progressive tax on consumption. According to him, citizens should pay taxes at a marginal rate that is positively correlated to their amount of consumption: the more they consume, the higher the marginal rate should be.

His proposition is rooted in the observation that individuals are trapped in a positional arms race, which produces externalities (i.e. costs imposed on third parties to an original exchange). In the United States and elsewhere3, individuals consume too much in general and too much of positional goods in particular, i.e. goods that are mostly demanded for their extrinsic qualities (i.e. for signalling one’ status).4

Frank’s central claim is that consumption is context-sensitive: other individuals’ decisions matter in one’s consumption. Individuals are influenced by social standards of spending, i.e. the rules about appropriate consumption. This context-sensitivity cannot be reduced to status-seeking alone as some suggest (Botton 2004) since someone’s relative standing may deeply affect his ability to pursue his life projects and realize his conception of the good (Frank 1999, pp. 122–145). People care about their relative standing (determined by their relative spending) because it affects their opportunities in life and the opportunities of people for whom they care (e.g. children, spouse, dependants).

The origin of the problem lies in the positional dynamics that have been accentuated by the dramatic increase of top-incomes since the 1970’s whereas low and middle incomes have stagnated or only slightly increased (Frank 2007, pp. 7–16).5 The crux of the argument is that top-earners strongly influence social standards on appropriate spending in relation to home, clothes, travel, and so forth. ‘Expenditure cascades’ (Frank 2011, p. 61) explain how these patterns spread throughout society: top-earners’ habits are imitated on a smaller scale by the income group who is below, whose consumption is imitated by the income group just below, and so forth.

By their spending habits individuals create positional externalities for others: for decades, standards for a “decent” home, gifts, clothes, vacation, cars, electronic appliances, and other goods and services have been rising. For instance, according to the National Association of Home Builders, the average size of US houses had almost doubled between 1970 and 2009. So, individuals have been using more resources for keeping up with criteria of socially adequate spending. Individuals might spend too much, to their own detriment. In addition, positional consumption creates prisoner dilemmas where agents are trapped in suboptimal situations.

These situations are suboptimal due to ‘positional externalities’: everyone’s spending increases the pressure on everyone else to adopt the same strategy.6 This pressure creates further damage: overwork, wasteful conspicuous spending, pressure to perform, psychological distress, etc. (Frank 1999). This damage results from ‘positional arm-races’ where most of participants try to outperform (outspend) their competitors. Frank sees these dynamics at work for housing, gifts, vacations, clothing, and so forth. As with the arms race between the Soviet Union and the United States during the Cold War (Brams 1985), the increased spending on housing or cars incurs an adjustment of the consumer’s immediate surroundings (relatives, friends, colleagues), which propagates further spending.7

Frank’s appeal to ‘positional externalities’ strengthens the case for public intervention, especially in relation to those who are reluctant to accept the state’s interference in private matters (Frank 2008, 2011, pp. 194–215). Based on John Stuart Mill’s harm principle 8, Frank argues that positional consumption generates a broad array of costs that justify public intervention because they are imposed on third parties. His argument is to add to the loss of welfare the constraints imposed by individuals on each other and the individuals’ incapacity to independently scale down their expenses (Frank 2008 p. 1782), which results in the impossibility of reducing the production of externalities in a significant manner.

Frank identifies two levels of positional harms. Firstly, individuals ruin their psychological situation. They do not optimize the happiness potential that could be derived from their material resources.9 They get less satisfaction or subjective well-being than they would get in at least one attainable alternative (i.e. Pareto improvement is possible).

Secondly, individuals ruin their material situation, both individually and collectively. Individually, expenditure cascades increase positional spending and activities whereas decreasing non-positional consumption and activities (e.g. time spent with friends and family). People save less and become more vulnerable to future adverse events (e.g. unemployment, illness). Less affluent individuals rely on bank credit for keeping up (and counter-balancing the stagnation of their income). Socially, private and public investments are squeezed, jeopardizing public goods and growth. Also, the increased pressure to “keep up with the Joneses” partly explains the current resistance to accept taxation.

As a solution, Frank proposes a progressive tax on consumption that will not apply on specific positional goods (contrary to a luxury tax), but on individuals’ volume of spending.10 Such a tax will curb positional consumption in particular and consumption in general. If the rates are steep enough, it will cause a decline in top-earners’ spending, but also in the spending of lower income classes in virtue of the expenditure cascades. Individuals will be forced to save more, which will be beneficial to private investments and enhance individuals’ capacity to cope with adverse events. Finally, more resources will be levied for financing public goods, which will benefit everyone in the society.

Frank’s argument is a mix of at least three different justifications, namely efficiency, paternalism and equality. The subsequent sections will present and discuss each of them.11 Then I will argue that the core of the issue is how to envisage social cooperation. My suggestion is that the choice of a tax base expresses a particular conception of social cooperation.

7.3 Efficiency

Usually economists and philosophers share the view that efficiency is an a-moral concept since it would be silent on the good and the bad, the right and the wrong features of different actions or states of the world. Thus, the efficiency of a social arrangement would be a purely empirical matter. This view is rooted in a division of intellectual labour between economists, who would deal with factual issues, and philosophers, who would deal with evaluative ones.

This division might have some cogency, but it cannot be denied that the concept of efficiency is partly normative. In order to judge that a given arrangement is efficient, two normative “things” should be decided: (1) what needs to be efficient, i.e. the object of efficiency, and (2) which principles must ground this evaluation, i.e. the rules of efficiency.


The debate on whether Gross Domestic Product (GDP) offers a fair evaluation of socioeconomic performance illustrates the first point (van den Bergh 2010). No doubt the choice between GDP or alternative indicators such as UNDP’s Human Development Index or the Happy Planet Index is to some extent influenced by practical considerations (e.g. the data’s availability, the possibility to sum different aggregates or undertake international comparisons). But the choice is normative too since it expresses views about what should be collectively valued (economic growth, education, life expectancy, environment, etc.).



Choosing between different conceptions of efficiency is not axiologically neutral either. For instance, a choice between Pareto and Kaldor-Hicks efficiencies is a choice between a conception that excludes the improvement of the situation of an agent at the expense of another (the so-called “minority sacrifice”)—Pareto’s conception—and a conception that precisely allows a trade-off as long as compensation is hypothetically possible—Kaldor-Hicks’ conception.


Efficiency is a normative concept that allows different interpretations (conceptions). Ultimately, the “what” of Frank’s conception (i.e. what should be efficient) boils down to the conditions of social cooperation. In his writings, there are at least three aspects in which implementing a progressive tax on consumption would render social cooperation more efficient: the reduction of positional externalities’ “leakage”, the maximization of the social output (especially through private and public investments), and the improvement of individuals’ satisfaction or happiness.

In that respect Frank is the heir of Welfare Economics that attributes to the state the responsibility of correcting ‘market failures’ (Baumol 1952, 1972), i.e. to enhance welfare. However his use of Mill’s harm principle expresses the additional argument that public intervention is ultimately justified by the prevention or correction of harms imposed on others (Frank 2008). Since market failures imply imposed harms to others, the state ought to intervene.12 Pigovian taxation is the classical tool. It consists of charging a tax on the emitters of the negative externalities (Landes 2013a). Frank’s originality is to stretch the concept of externalities and Pigovian taxation for including positional dimensions.

However, Frank does not clearly define the concept of positional externalities in his writings. A plausible interpretation is that a positional externality consists of the negative variation of the status of a given individual that results from an initial interaction of which he has not been part (implying that he has not agreed to the terms of this interaction), inducing two kinds of costs: material ones (e.g. high consumption, low savings, or under-financing of public goods like roads, bridges, hospitals, etc.) and psychological ones (e.g. reduced well-being and happiness).

By contrast, efficient cooperation is characterized by limited “positional leakage”, high private/public investments and high individual satisfaction/happiness. However, if the objective was to reduce or suppress positional externalities, why is a (incremental) tax on consumption more efficient than, say, a (incremental) tax on income or wealth?

It is one thing to justify the efficiency of a tax on consumption, it is another to justify its superior efficiency vis-à-vis a tax on income or wealth. Superior efficiency does not follow from efficiency. For instance, tax structures relating to income or wealth can become so quickly confiscatory that conspicuous consumption will be reduced in larger proportions than offered by any tax on consumption. Also, if the problem is positional externality, why not capping income and wealth through regulation? In short, why is taxing consumption is the preferable option?

In the literature, the principle of ‘neutrality’ or ‘non-discrimination’ between consumers and savers advocates for consumption as a more appropriate tax base than income.13 According to Mill, taxing income leads to a double taxation of the savers (Mill 1848, V, I, 4): first, when they earn their income, even if part of it is invested, and, afterwards, on the return of this investment.14 Beyond the ‘unfairness’ to agents in proper (Andrews 1974), i.e. the advantage given to some lifestyles over others, this absence of neutrality is detrimental to private and public savings and investment, i.e. efficiency.

The last point is the reason Frank emphasizes in support of substituting a tax on consumption in the current US tax system: reducing positional externalities without dis-incentivizing economic activities. If consumption is taxed at a steeply progressive rate, individuals will save more, which will positively affect investment and economic growth. Moreover, work will not be discouraged since the accumulation of wealth and investment will be rewarded.

Efficiency, then, justifies taxing consumption, instead of income, since it helps to tame positional externalities while not dis-incentivizing agents to engage in productive activities and, furthermore, incentivizing them to invest. However, Frank’s conception of efficiency, as well as common defences of consumption tax, appears to include more than a consequentialist calculation. A “thicker” normative ideal seems to inspire the choice of consumption as the appropriate tax base arising from the idea that consumption is wasteful, especially in affluent societies. Furthermore, consumption is presented as being the source or vehicle of perverse social interactions implying signalling and counter-signalling, status-seeking, rat-races, etc. (James 2007; Schor 1998).

The taxation of consumption is, at least partly, driven by an underlying moral commitment to anti-consumerism or by an ascetic ideal according to which consumption should be directed towards the satisfaction of basic needs, the rest often being superfluous or harmful. This commitment would explain the heavy emphasis on non-positional goods such as family, social relations, environmental concerns, altruism and generosity. Such emphasis would not be fully explainable by efficiency alone, but by a deeper conception of what has value in life.15

Only gold members can continue reading. Log In or Register to continue