Isolationism, instrumentalism and fiscal policy

Abstract

When reading contemporary theories of distributive justice, one could easily get the impression that questions of fiscal design are normatively speaking merely instrumental for realizing the distributive ideal. Once the overall conception of justice is settled upon, questions of how the state should arrange its institutions and policies are settled if they effectively and efficiently promote the preferred distribution. I argue that such pure instrumentalism is mistaken in the context of fiscal policy. As a result, there is nothing problematic or morally arbitrary about accepting domain-specific principles of fiscal justice.

1. Introduction

When reading contemporary theories of distributive justice, one could easily get the impression that questions of institutional design are normatively speaking merely instrumental for realizing the distributive ideal. Once the overall conception of justice is settled upon, questions of how the state should arrange its institutions and policies are settled if they effectively and efficiently promote the preferred distribution. Here is, for example, John Rawls in his famous section on the subject of justice:

A conception of social justice … is to be regarded as providing in the first instance a standard whereby the distributive aspects of the basic structure of society are to be assessed. (Rawls 1971: 9)

In this passage, Rawls talks about the basic structure in its entirety. That is, you cannot assess the justice of one institution that is part of the basic structure without considering how it interacts with others. ¹ The question, for example, whether or not to subsidize daycare for young children has to be decided by looking at whether such a policy would contribute to an overall distribution of rights, duties, burdens other advantages that satisfies the principles of justice. Should such a subsidy turn out to be mostly advantageous to middle-class women and not to the least well off given how the rest of society is organized, then the principles of justice counsel against such subsidies and urge, for example, means-tested income support instead. Call the view that the choice of a particular policy and institutional design is only to be evaluated in terms of its contribution to the overall pattern of distribution instrumentalism. Instrumentalism holds that once you have settled upon an over-arching conception of distributive justice, there remain only pragmatic questions about the right policy and institutional design. As a result, instrumentalism denies that there are domain-specific principles of justice.

Instrumentalism is particularly dominant in discussions about fiscal policy. Whether or not we should tax capital income differently from other sources of income; whether or not mortgage interests should be deductible; whether there should be an exemption from income taxation for the lowest incomes – all such questions are to be decided by figuring out the overall effects of such measures combined with other activities of the state (such as subsidies and regulation) on the preferred distributional pattern. A good example of such an instrumentalist approach can be found in Liam Murphy and Tom Nagel’s The Myth of Ownership:

The real issue of political morality is the extent to which social outcomes are just, and knowledge of the distribution of real tax burdens is important only insofar as it helps us advance that aim. (Murphy and Nagel 2002: 131)

The idea that there might be principled moral constraints on the distribution of tax burdens independent of the overall distribution of duties, rights and advantages, or that there are special principles of fiscal justice is a mistake on this view.

It did not use to be this way. Older generations of tax lawyers and economists, prior to WWII, would start their handbooks with a discussion of the legitimacy of taxation as such. ² Could a state legitimately require citizens to pay for its activities? The answers proposed would then lead the authors to conclude that the state is only justified in levying taxes if this is done in accordance with some basic domain-specific principles of fiscal justice, such as the benefit principle (‘from each according to how much they benefit from the state’) and – especially – the ability-to-pay principle (‘from each according to their ability to pay’).

Both these principles determine the fair share of tax burden by taking the pre-tax distribution as given. ³ For example, the benefit principle determines the amount of benefits received from state activity by looking at the pre-tax distribution. Then, assuming that the appropriate tax base is income, only one’s pre-tax income, or rather, only that part of one’s pre-tax income that is the result of state activity, is the morally relevant departure point for determining one’s fair share in income taxes. Similarly, the ability-to-pay principle looks at some pre-tax feature (e.g. one’s pre-tax income) to determine what a person’s ‘ability’ is and what would be a fair share, given that ‘ability’. ⁴

The resulting view is that questions about the fairness or justice of taxation tended to be evaluated in isolation of questions about the overall distribution of income, benefits and opportunities. A striking example of this way of thinking can be found in an old textbook on the Dutch fiscal system in which the author defends an ideal of tax neutrality:

In a sea of injustice in the area of income distribution, only those taxes are just that limit themselves to their task: to finance the productive side of government policy, about which it has no judgment, and to distribute the burdens evenly. (Bordewijk 1930: 337) [translation by me]

Let’s call such views isolationist. Unlike instrumentalists, isolationists do accept special domain-specific principles of justice. Moreover, they believe that one should determine the justice of a particular institution independently of other institutions.

Pure isolationism and pure instrumentalism are the opposite ends of a spectrum in that each explicitly denies what the other endorses. Where pure isolationism denies the relevance of overall distributive concerns for the normative evaluation of fiscal policy, pure instrumentalism claims that these concerns are the only ones that matter.

In this paper, I argue that pure instrumentalism is mistaken and that questions of institutional design are not merely instrumental for the just overall distribution of social outcomes – at least in the domain of fiscal justice. As a result, there is nothing problematic or morally arbitrary about accepting special domain-specific principles of justice – again, at least in the domain of fiscal justice. This does not mean, however, that the old isolationists were right after all. Against pure isolationist views like that of the quote above, I will argue that considerations of the overall distribution of social outcomes can, and should, play a role in the assessment of fiscal policy. What is more, close inspection of the ability-to-pay principle and the benefit principle show that these principles are compatible with such distributive concerns and to some extent presuppose them.

The argumentative strategy that I will follow is the following. First, I will discuss four objections one can find in the literature against domain-specific principles of fiscal justice. The first three are the productive factors objection; the myopia objection; the conventionality objection. These three objections attack domain-specific fiscal principles indirectly by denying any moral relevance to the pre-tax distribution. Since domain-specific principles – such as the benefit principle and the ability-to-pay principle – determine fiscal duties by appealing to that distribution, this shows, so these arguments claim, the arbitrariness and even incoherence of domain-specific principles of fiscal justice. I then proceed to formulate various replies that show that such principles are not arbitrary or incoherent.

The fourth objection is the public-private division objection. It is different in that it attacks domain-specific principles of fiscal justice directly. Such principles, so the objection goes, are silent on the question how much tax should be raised. That is, how much of society’s resources should be left in private control of individuals and how much should be under public control of the government is not answered by domain-specific fiscal principles. I then answer this objection by showing that these principles do have implications for how fiscal revenues ought to be spent and, consequently, are far from silent on the public-private division.

Finally, having answered the objections against domain-specific principles of taxation, I will argue that there are substantive domain-specific normative constraints of fiscal justice. The result is a view on fiscal justice that on the one hand accepts the concerns of overall distributive justice but insists on some moral constraints in the domain of taxation.

2. Productive Factors

The first objection results from an obvious point. Suppose persons A and B have the same pre-tax income from their respective occupations and are similar in all other respects. Suppose both pay the same amount of tax. Suppose both A and B decide to insulate their homes. A receives a (non-fiscal) subsidy from the government for this. B, on the other hand does not receive such a subsidy. On the pure isolationist view outlined above, A and B have been burdened fairly by the tax system: taxation has ‘left them as it found them’. However, the difference in disposable income between A and B seems unfair. Assessing the fiscal treatment of A and B cannot be independent of the fact that A receives that subsidy and B does not. More generally, it seems wrong to discuss tax burdens without considering what is done with the revenues of taxation. Therefore, if one insists on special principles of justice in taxation, they had better be such that public expenditures play a role in them.

The productive factors objection against pure isolationism says that the problem with isolationist domain-specific principles is deeper than this. Such principles treat the pre-tax distribution as the relevant point of departure for assessing tax burdens. So even if you amend the type of pure isolationism of the pre-WWII tax lawyers by including some evaluation of public expenditures in the principles, you still need to demonstrate that you are correct in treating the pre-tax distribution as morally salient, as the right point of departure for moral evaluation. As mentioned above, instrumentalists think this a mistake. The productive factors objection is an attempt at explaining why that is.

The pre-tax distribution is the result of the choices and transactions of all individuals. These choices and transactions do not occur in some state of nature. They are structured and made possible against the background of all kinds of legal, social and conventional structures that are the result of state activity. For example, a simple economic transaction between two individuals is only possible if these individuals have property rights, including the right of transfer, over the transacted goods, as well as the assurance that these rights are protected. State activity is crucially important for just about any transaction. The state provides the infrastructure over which goods are transported from supplier to customer; goods are manufactured by workers, educated (in part) by the state; these goods are safe in the factory and the warehouses, because the state provided protection in the form of police and fire forces, etc.

Once this is clear, so the argument goes, it is obvious that the pre-tax distribution itself is the result of massive state activity. Without that activity, there would not be a pre-tax distribution, for we would all be in a state of nature in which there would be no property, no industry, and no safety. Therefore, the pre-tax distribution itself is already the result of all kinds of public expenditures. Instrumentalists conclude from this that the pre-tax distribution is morally irrelevant. People do not have any claim to their pre-tax holdings. Only their after-tax holdings are theirs in any morally relevant sense. ⁵ It is clear that if this is the correct way to think of the pre-tax distribution, isolationism, which takes as its departure the pre-tax distribution, is mistaken. Let’s call this the productive factors argument, as it emphasizes the fundamental role that state activity has for the products of social cooperation.

As it stands, the productive factors argument is fallacious. From the fact that without the state activities, we would all be living in a state of nature, it does not follow that therefore the entire pre-tax distribution is due to the state. From the fact that without fertile soil a farmer could never get any harvest it does not follow that therefore the entire harvest is due to the soil. After all, the effort, talent and experience of the farmer are also necessary. Of course, it does not follow either that only the effort, talent and experience of the farmer are responsible for the value of the harvest as Locke claimed. ⁶ Both factors – soil and labour – are crucial in producing the harvest. Similarly for the role of the state in the production of the pre-tax distribution: without the efforts, ambitions and talents of people, there would not be any pre-tax distribution either. Similarly, it does not follow that, therefore, people do have an exclusive, absolute right to their pre-tax shares, as many libertarians hold. ⁷ Instead, no interesting conclusions about the moral (ir)relevance of the pre-tax distribution follow from the fact that without the state’s activities there would not be a pre-tax distribution. ⁸

3. Myopia and Conventionality

Liam Murphy and Thomas Nagel use a subtly different argument for their version of instrumentalism. They argue that all isolationists subscribe to what they call ‘everyday libertarianism’, which holds that taxation takes from individuals what is legitimately theirs. ⁹ Any tax will infringe upon this property right and infringements on such basic rights need a special justification that recognizes this right. Supposedly, the special principles of fiscal justice, such as the ability-to-pay principle, do just that. They take as their point of departure the pre-tax income distribution and then spell out the conditions under which the state is justified in moving away from that distribution by violating that property right.

However, everyday libertarianism, Murphy and Nagel claim, is arbitrary and incoherent. It is arbitrary because it takes part of the state activities as given while assessing the justice of other parts, without any good reason as to why those activities (such as the rules of the market) are taken as given and others (such as taxation) are not. ¹⁰ And it is incoherent because one should take either all state activity as given before giving a moral assessment of the resulting distribution, or one should take none of the activity as given before assessing the moral justification of the state’s activities, much like philosophical libertarians do. Everyday libertarianism, then, is guilty of a kind of intellectual shortsightedness. This is the myopia argument against isolationism. ¹¹

Murphy and Nagel have a related argument against isolationist special principles of fiscal justice. They argue that property rights are fundamentally conventional: they are a legal fiction. In other words, property rights are granted by the community. Since the community also has conventions regarding taxation, there is nothing morally exceptional about the pre-tax distribution. There is not a deep moral sense in which it is ‘mine’ or ‘yours’. Moral evaluation must decide how what is ‘mine’ and what is ‘yours’ should be determined. The conventions of property, markets and taxation need to be critically assessed to see which set of conventions best serves the aims of a just society. Therefore, that assessment cannot appeal to property rights, let alone property rights to pre-tax shares, for it is the very system of property that is under scrutiny when we make such assessments. ¹² Call this the conventionality argument against isolationism.

In formulating these two arguments – myopia and conventionality – Murphy and Nagel clearly think that isolationism is the result of some kind of everyday libertarianism, that takes the moral relevance of the pre-tax distribution to be that of a (moral) property right.

Isolationists could defend themselves against the arguments of Murphy and Nagel in two ways. First, by denying that isolationism is committed to a kind of everyday libertarianism; secondly, by rejecting both the myopia and conventionality arguments. I discuss these in the next two sections.

4. Isolationism and Everyday Libertarianism

Let’s start with the first way. Isolationists could argue that their preferred principles of just taxation are not committed to everyday libertarianism. That is, they do not need to assume that people have a property claim to their pre-tax shares in order to make sense of their principles. Consider, for example, a typical application of the ability-to-pay approach to income tax. All states that I am familiar with, exempt an initial amount of income from taxation. Let’s say this amount is €8000. Any income above that is then taxed, either at a fixed, flat rate or progressively. Why is this amount exempted? If everyday libertarianism is what motivates the ability-to-pay principle, it must be because the claim of property to the first €8000 is stronger than the property right to the rest of one’s income. However, this is unnecessary to defend such exemptions. For example, defenders of the ability-to-pay principle can (and do!) argue that below €8000 individuals have no ability to pay any taxes and therefore should be exempt. ¹³ This justification is the straightforward implication of the ability-to-pay principle and not the result of some exceptionally strong and overriding property claim on the first €8000 of one’s before-tax income. Similarly, a defender of the benefit principle, which holds that people should pay taxes to the extent that they benefit from the state’s activities, can justify this exemption by arguing that if somebody earns less than this amount, she is not really benefitting from the state, or not benefitting sufficiently. Again, this justification does not rely on any property claim to pre-tax distributions. In other words, everyday libertarianism, the view that people have property rights in a share of the pre-tax distribution, is not essential for typical isolationist justifications of tax burdens.

It is clear that for isolationists it is essential that the pre-tax distribution have some moral relevance in order to explain why it is the proper baseline from which to assess tax burdens. However, this moral relevance need not involve anything as strong and strict as a property right. So, while Murphy and Nagel’s claim that isolationists are necessarily guilty of an implausible everyday libertarianism is misplaced, it signals an important question for isolationists: what is the moral relevance of the pre-tax distribution? Therefore, this first line of defence against the critique of Murphy and Nagel is not entirely successful.

5. Rejecting the Conventionality and Myopia Objections

This brings us to the second way of defusing the arguments of Murphy and Nagel: rejecting the myopia and conventionality objections. I start with the second one, the conventionality objection, because it has attracted most attention from commentators. ¹⁴

Suppose that one accepts that property is intrinsically conventional and that one grants that fiscal policy is conventional as well in the way that Murphy and Nagel identify; does this mean that the pre-tax distribution is an amorally irrelevant point of departure for assessing the acceptability of tax burdens? It does not. The conventionality of a set of rules does not mean that they are, therefore, morally irrelevant or arbitrary. ¹⁵ First of all because the fact that a rule is a convention only says something about its origin. The claim that the rules of private property are conventional, just means that they are the product of past human artifice. But the genealogy of a rule does not tell us anything about whether it is morally relevant. And this brings me to the second point: conventional rules can be as relevant as non-conventional moral principles. ¹⁶ Just because driving on the right in the USA is a convention, it does not follow that it is morally indifferent or arbitrary to drive on the right or the left in the USA. Similarly, if ordinary people conventionally attribute moral relevance to the pre-tax distribution in their judgements about the fairness of fiscal policy, the pre-tax distribution need not be morally irrelevant or arbitrary because of that. In fact, it would make the pre-tax distribution a salient starting point for the evaluation of fiscal rules. Therefore, the conventionality argument does not compel one to reject isolationism, nor does it lead to instrumentalism.

More importantly, the conventionality argument is not the best argument for instrumentalism, for it depends on a controversial claim about the nature of property – namely, that property rights are intrinsically conventional. But many critics of Murphy and Nagel explicitly deny this assumption. As far as I can see, Murphy and Nagel do not offer any explicit argument for their claim about the conventional nature of property. Instead, they see a deep connection between taxation and the institution of property:

The conviction that determines our approach to all more specific questions is that there are no property rights antecedent to the tax structure. Property rights are the product of a set of laws and conventions, of which the tax system forms a part. (Murphy and Nagel 2002: 74)

However, this claim about the connection between taxation and property does not imply the conventional nature of property. One can insist on the non-conventional nature of property and evaluate the laws and conventions, including those of the tax system, on that basis precisely because there is this deep connection. In fact, that is exactly what the everyday libertarian is doing. ¹⁷

I conclude that the conventionality argument does not show that the pre-tax distribution is the wrong place to start from when evaluating the tax system. Instead, the myopia argument is more promising for the critics of isolationism, as it does not seem to rest on assumptions about the nature of property rights that isolationists can simply reject.

Govert den Hartogh has argued that everyday libertarian isolationists, as well as Murphy and Nagel when wielding the myopia argument, share an assumption regarding the nature of property rights. ¹⁸ Everyday libertarians regard their share in the pre-tax distribution as their property. Any claim on this property by the state is a potential violation of that right. Instrumentalists such as Murphy and Nagel argue on the contrary that one’s share in the pre-tax distribution is not one’s property – only one’s post-tax share is – precisely because the state still has claims on that share. Therefore, both everyday libertarians and instrumentalists assume that a property right to something cannot be burdened with positive duties. ¹⁹ Everyday libertarians use this to decry taxation as a form of theft; instrumentalists use this to argue that, therefore, there is no property right in one’s pre-tax holdings in the first place. Therefore, both everyday libertarians and instrumentalists assume that property rights to something are incompatible with additional positive claims of others to that something.

This is a questionable assumption. For example, suppose that I take out a mortgage from the bank to buy a house. The house is now my property. I also have a duty to make monthly mortgage payments to the bank and if I fail to do so, the bank can seize my house. But this does not mean that I don’t have full property to my house. Nor does it mean that the bank really owns my house and that I am merely renting from the bank. Home ownership is compatible with there being all kinds of obligations with regards to that home. Similarly, the fact that I have additional duties to pay taxes over my pre-tax income does not imply that this income is not really mine and that I am just the administrative placeholder where this part of the social product is temporarily parked.

If this makes sense, then this reason to believe that isolationists are guilty of myopia is removed, for it is perfectly coherent to accept that one has duties, positive duties, over one’s property rights. ²⁰ However, we have not completely answered the core of the myopia objection. We still need to demonstrate that applying domain-specific principles of fiscal justice is not arbitrary in taking the pre-tax distribution as the morally relevant point of departure. In the next section, where I discuss the public-private division objection, I will show that taking the pre-tax distribution as the starting point for assessing fiscal burdens is not arbitrary or incoherent.

6. The Public-Private Division

If the alleged arbitrariness and incoherence of special principles of fiscal justice is not due to any form of everyday libertarianism, are there other reasons to believe they are arbitrary or incoherent? I hinted at such a reason in the initial discussion of the problems facing pure isolationism. Pure isolationism disregards completely what is done with the revenues of taxation, though this clearly matters for the assessment of the tax burdens. Murphy and Nagel use this observation to claim that all traditional principles of fiscal justice necessarily are incomplete, as they do not tell us how much tax is to be collected and how it should be spent. ²¹ That is, traditional principles such as the ability to pay principle and the benefit principle give no answer to ‘how much of society’s resources will come under the control of government, for expenditure in accordance with some collective procedure, and how much will be left in the discretionary control of private individuals, as their personal property’. ²² They call this the public-private division.

The objection against special principles of fiscal justice then is that their silence on the public-private division renders them implausible since the public-private division is crucial in determining the total tax burden and its distribution. It is not entirely clear why silence on the public-private division necessarily renders special principles of fiscal justice implausible. After all, there can be general principles of distributive justice that answer this question and the special principles may be crucial in the decision of how to distribute the resulting burdens.

Be that as it may, Murphy and Nagel’s claim that traditional principles cannot answer the question to the public-private division is surprising given their own solution to the public-private division. ²³ There is no need to go into the minutiae of their proposal; a rough sketch is sufficient. Imagine a society where the market generates just outcomes, so that the pre-tax distribution is just. The optimal and fair level of taxation and public expenditure is then to be calculated as follows: every subject should be levied with that amount of tax for which the marginal utility of their private spending is equal to the marginal utility of their public spending. If the market generates different pre-tax holdings or if people’s utility schedules differ in their preference for public versus private expenditure, the result will be a differentiated rate structure. ²⁴ Obviously, it is very hard to determine if, and if so, how much, a specific level of public activity by the state or a specific level of private spending is worth to an individual, let alone the marginal utility of public and private spending. ²⁵ However, the point is that Murphy and Nagel claim that there is a principled answer to the question as to what is a fair and just division of public and private spending, and the criticism of Murphy and Nagel then is that special principles of fiscal justice cannot generate that answer.

The irony is that Murphy and Nagel’s solution to the public-private division is completely in line with a traditional “myopic” approach to taxation. It is essentially the solution that Erik Lindahl offered to this question, and it is a version of the benefit principle. ²⁶ In other words, here is a traditional principle that does give prescriptions for public-private division. ²⁷ And what is more, Murphy and Nagel explicitly endorse it!

However, Murphy and Nagel maintain, their solution only works if the pre-tax distribution is just, which means that all claims of (re-)distributive justice have been met. That is, in their view the approach can only say something about the financing and the provision of public goods and services once the ‘business of distributive justice’ is done. ²⁸ And since the content of this ‘business’ for a large part depends on the public-private division, they consider the exercise completely hypothetical. As I explain below, this is not correct.

Here is how a defender of traditional domain-specific principles of fiscal justice could answer the public-private division argument. Consider, first, the purposes of taxation. The traditional theory of public finance says that taxation is used (1) to finance the production of public goods and services by the state; (2) to implement redistributive policies; (3) to conduct economic policy (e.g. reducing sovereign debt or controlling inflation). ²⁹ Murphy and Nagel add a fourth category, namely (4) the discharge of public duties. This last category consists of things like the provision of development aid, or support in cases of emergencies (like an earthquake or fire). ³⁰ Such activities are not public goods or services strictly speaking: they are exclusive (groups and individuals can very well be excluded from benefitting from the services of the fire department) and rivalrous (every time the fire department is extinguishing a fire, they have diminished capacity to respond to other emergencies). Public duties are duties that each individual owes to the victims of such events. However, due to the nature of these calamities, it would be ineffective and inefficient if each sought a private way of meeting these duties. (Imagine what would happen after a serious earthquake if all citizens rush to the rescue of the victims in the affected area.) Instead, we see to it that these duties are met through the activities of the state: emergency services, trained and ready to respond immediately, financed through taxation. ³¹

Murphy and Nagel argue that their favourite solution to the public-private division is applicable to (1) the provision of public goods and services and (4) the discharge of public duties – but not to (2) the implementation of redistributive policies and (3) specific economic policy. Since these latter functions of taxation influence the pre-tax distribution, the solution to the public-private division, so they argue, is completely hypothetical.

It is unclear why (2) and (3) should be excluded from the solution. I will start with (2), the implementation of redistributive policies. We can do the same for these policies as we have done for the discharge of public duties. Public duties are to be financed through public means if this is the most effective and efficient way to discharge those duties. The ideal level of expenditure for these duties can be determined in the same way as public goods and services can be determined, provided the pre-tax distribution is just. Duties of justice are individual duties that, just like public duties for assistance with emergencies, are most effectively discharged through collective action. Therefore, we can use the pre-tax distribution as our starting point for determining which goods and services and which redistributive policies should be pursued, just as we could use the pre-tax distribution to determine how and at what level we should finance the collective discharge of public duties. What is more, we can use versions of the traditional principles of fiscal justice (e.g. the benefit principle as developed by Lindahl and supported by Murphy and Nagel) to determine both the optimal level of these and the corresponding distribution of tax burdens. ³² In other words, since the (level of) financing of public duties can be justified by an appeal to traditional principles of fiscal justice, so can the financing of other duties of distributive justice. Just like Murphy and Nagel’s category of public duties, the duties of distributive justice are individual duties that are most effectively collectively discharged. Therefore, one does not have to presume that the pre-tax distribution is just when one approaches the question of the distribution of tax burdens using traditional principles of fiscal justice that give moral relevance to the pre-tax distribution. Just as the collective discharge of public duties can be justified by an appeal to a domain-specific principle of fiscal justice, so can the collective discharge of duties of distributive justice. None of this begs the question against the moral relevance of the pre-tax distribution, nor is the approach in any way arbitrary or incoherent.

7. Taxation and Economic Policy

That leaves the third purpose of fiscal policy mentioned above, the specific economic policies that are often pursued through taxation. It follows from the argument so far that if specific economic policies (3) for which fiscal instruments are used can be justified as a matter of public duty or as a duty of redistributive justice these too can be included.

However, it should be recognized that this is not always the case. This does not mean that such policies cannot be justified – of course they can. The crucial point is that often the goals of such policies do not constitute a duty for individual taxpayers whereas public duties and duties of redistribution do. There may be good reasons for states to pursue certain economic goals, such as limiting sovereign debt or controlling levels of inflation. Furthermore, it certainly is correct to say that individual citizens may benefit from the successful pursuit of such goals. Limited sovereign debt and realizing relatively low levels of inflation are conducive to the economic opportunities and welfare of citizens. What is more, it could be the case that individual citizens have reasons to cooperate with the state in the pursuit of these goals. Perhaps the reasons supporting such goals outweigh other reasons and may justify the state to compel citizens to behave accordingly. However, it is one thing to claim that there are good reasons for the state to pursue certain goals but quite another to claim that individual citizens have an individual duty to pursue those goals on a par with the duty to help victims of catastrophes or injustices. Because an individual’s reasons to pursue economic policies differ from the reasons to provide public goods and services and contribute to collective ways of discharging individual duties of assistance (i.e. public duties and duties of justice), we cannot use the same argumentative strategy for this category of purposes of fiscal policy. Critics of all too enthusiastic use of fiscal policy in the service of macroeconomic goals, will welcome this conclusion, as it shows that policymakers cannot rely on the same principles that inform taxpayer compliance in the case of the other purposes of taxation distinguished above. They will extrapolate and claim that taxation in the pursuit of justifiable economic policies cannot be justified. ³³ Obviously, that conclusion is too strong. All this argument shows is that a plausible justification of the use of fiscal policy in the pursuit of economic policy will have to be a different one than the one that justifies the use of taxation for the financing of public goods and services or the discharge of public duties and duties of distributive justice. ³⁴

Be that as it may, the crucial point of this exercise is that the charge that domain-specific principles of fiscal justice are silent on the public-private division as Murphy and Nagel claim is misplaced. The benefit principle, for example, is far from silent on which kind of state activities should be pursued through fiscal means.

8. Constraints on Pure Instrumentalism

So far, I have shown that the choice between isolationist and instrumentalist approaches is not an all-or-nothing affair; a choice between an implausibly arbitrary and restrictive myopia or a purely outcome-oriented approach to evaluating fiscal policy that denies the existence of any special moral restriction on the use of the tax system in the pursuit of justice. In particular, I hope to have convinced the reader that one can be concerned with redistributive considerations on the one hand while treating the pre-tax distribution as the morally relevant baseline for evaluating tax burdens on the other. ³⁵

It might seem then that there are two plausible approaches to fiscal justice: on the one hand, an instrumentalist approach that is concerned with overall outcomes and, on the other, an approach that uses special principles of fiscal justice that take the pre-tax distribution as the morally relevant starting point. It would then be a matter of taste or expediency which of these one adopts in assessing fiscal policy. However, that would be a mistake. In what follows, I will outline some restrictions of fiscal justice that mitigate a pure instrumentalist approach to taxation. Just as pure isolationism should be rejected, a pure instrumentalism that only is concerned with outcomes should be abandoned.

Consider, first, some feasibility constraints on the use of fiscal instruments in the service of some policy that any government will have to consider. Often it is not expedient to use tax facilities to promote social goals since other non-fiscal instruments (most notably, subsidies) are more effective and efficient and have fewer unintended side-effects. ³⁶ In general, fiscal measures are relatively blunt instruments in comparison to targeted subsidies and other forms of state interventions. Furthermore, instrumentalist fiscal policies tend to lead to complicated tax laws with many exemptions and deductions, which enable subjects to game the system and avoid taxes. ³⁷ These sorts of circumstances are precisely what a smart instrumentalist approach will have to consider. This does not militate against an instrumentalist approach.

Another kind of feasibility constraint that any instrumentalist approach will have to contend with points to certain substantive and moral constraints on fiscal instruments. The literature on the psychology of taxpayers shows that most people see their taxes as a contribution from their pre-tax holdings to the state. ³⁸ They are willing to contribute, provided (1) they perceive their tax burden as fair; ³⁹ (2) that they believe that the revenues are spent effectively and efficiently on legitimate policies roughly speaking, ⁴⁰ (3) that they are approached by tax authorities as contributors, not as mere ‘subjects’ or convenient ‘locations’ from which to collect revenues; ⁴¹ and – and this is important – (4) provided they expect others to contribute their fair share as well. ⁴² In other words, their willingness to contribute is conditional. Since it is impossible for a government to audit every tax return, let alone prosecute each and every violation, governments have to rely for a large part on the voluntary compliance of citizens with the tax code. ⁴³ This puts some further feasibility constraints on pure instrumentalist fiscal policy. Governments must ensure that their fiscal instruments live up to these conditions or be prepared to deal with the consequences of large-scale declining tax morale.

All four conditions refer to the expectations and beliefs of subjects about the tax regime. Governments must engage in all kinds of activities to make sure that these expectations are met. So, governments have to accept some further constraints on pure instrumentalism if only for instrumental reasons. These constraints are both procedural and moral.

Procedural constraints have to do with familiar principles of legislation and the rule of law. An ideally instrumental income tax, for example, sometimes would announce the tax to be paid afterwards, so as to avoid that subjects can respond by changing their behaviour and then disrupt the outcomes aimed for. That is to say, purely instrumental taxation would have no principled objection to retroactive legislation. ⁴⁴ Similarly, sometimes it would be instrumentally ideal if tax policy would not be publicly announced. For example, if a company gets a certain favourable tax ruling, say, permission to use certain depreciation schedules on their equipment in order to enable the company to hire more employees, making this public could have the result that others start to apply for a similar permission as well, thus undermining other policy goals. That would counsel against publicizing such rulings, violating the ideal that law is public. Finally, an instrumental approach to taxation would not be averse to complex tax rules, which lead to fiscal ignorance, if the policy goals are achieved. ⁴⁵

If the four psychological conditions about tax morale outlined above are empirically plausible – and I have no reason to believe they are not – then retroactive, secret and complex fiscal instruments are to be regarded critically even for instrumental reasons, for they would undermine the willingness of people to comply. Retroactive policy, for example, would undermine the perception that taxpayers are regarded as contributors to public goods. Secret tax measures would undercut the perception that others are paying their fair share. That is, if you know that there are such secret exceptions, it is perfectly rational to expect that the average share that others have in the tax burden diminishes, which in turn leads one to decreases in tax morale. Finally, complex instruments leading to fiscal ignorance have the general effect that people are unaware of the true cost of the state activities thus increasing the belief that one ‘pays too much’, which reduces the perception that the government is efficient in its activities. ⁴⁶ This list of procedural constraints is not meant to be exhaustive.

The considerations above demonstrate that a purely instrumental approach to fiscal policy is not feasible given the specifics of tax motivation. It could be argued that none of this shows that an instrumental approach is fundamentally mistaken – just as with all goal-directed activity, external factors co-determine the range of feasible options. However, the feasibility constraints that the psychology of taxpayers poses on the fiscal policy point to underlying moral constraints and ideals. The government in issuing its fiscal measures must be a reliable, trustworthy party. It has to earn and maintain the trust of citizens – and it has to put some trust in citizens.

There are several ways of arguing for this claim. My favourite explanation invokes insights from the conventionalist theory of norms. A norm, whether legal, informal, conventional or otherwise, consists of a stable network of mutually reinforcing and stabilizing expectations that guide the actions of those subject to it according to the conventionalists.

A simple example can illustrate the basic idea. Suppose you and I approach a crossroads at the same time. I come from the left; you come from the right. We both want to get to our destination as quickly and safely as possible. What should I do? If I expect that you will maintain speed, I will have a reason to give way. That way, I avoid an accident and I will get home in one piece. So my reason to give way depends on my expectation that you will maintain speed. If you expect me to give way, you have a reason to maintain speed. In other words, our reasons, based on our expectations, refer to each other. They are interdependent. Suppose that most drivers in our community will maintain speed when coming from the right and give way when coming from the left. Then I will expect, with some confidence that you will maintain speed. I then will give way and you, following similar reasoning will maintain speed. In doing so, we reinforce the general expectation that cars from the right will maintain speed. ⁴⁷

Such general expectations tend to become moralized. Violations are met with resentment and indignation because we expect a certain amount of goodwill of others. If you maintain speed while coming from the left and we have an accident, I will blame you: I will believe that your action is the result of inexcusable ignorance or an inexcusable lack of goodwill towards me. What does that goodwill consist in? It consists of a basic attitude that you will not fail to live up to my reasonable and legitimate expectations even if that, here and now, in this situation, happens to be in your interest. For example, imagine you happen to drive a tank and I ride a brittle bicycle. Then you have nothing to fear or lose from a collision with me. Still, it would show a fundamental lack of goodwill to me if you were to maintain speed. I will expect that you won’t let me down; I will expect that I can rely on you – that is, I will presume a basic attitude of trustworthiness in you, for that is why I trust you. In other words, the rule in our community to give priority to cars coming from the right – even though conventional, as the rule could equally well have been different – presumes for its very existence a basic attitude of trustworthiness among participants. If this attitude is absent (or perceived to be absent), the norm is likely to collapse as it only relies on the perceived self-interest of traffic participants.

Obviously, this mundane traffic rule is simple in comparison to the complex regulations of the tax code. However, tax rules too can be analysed in terms of more or less robust patterns of interdependent expectations. ⁴⁸ Therefore, here too we see that if citizens are to have a reason to comply with the tax code, it cannot be pure self-interest – and the empirical evidence seems to confirm this. A government that wants to tax subjects will have to trust them and in return be trustworthy. It will have to live up to the reasonable and legitimate expectations citizens form about their tax burdens. It will have to ensure that the distribution of these burdens is fair. In other words, the tax system will have to display some basic moral institutional virtues such as trustworthiness and fairness.

A tax system that observes the type of constraints outlined above (both procedural and substantive) is more trustworthy and fairer than a system that does not. In other words, if the conventionalist theory of norms is plausible – and I believe it is – then there are important institutional virtues that the fiscal system will have to display. It will have to avoid retroactive, secret and complex fiscal instruments for these are detrimental to the trustworthiness of the government. Similarly, it will have to avoid regressive tax rates and, more generally, observe the ability to pay to promote the fairness of the system.

9. Conclusion

In this paper I discussed four arguments against special principles of fiscal justice. I showed that taking the pre-tax distribution as the morally relevant baseline need not be guilty of arbitrary and incoherent myopia. I also argued that there are internal domain-specific constraints on fiscal policy that substantially limit the use of taxation for achieving specific economic goals. The resulting view is a compromise between pure isolationism and pure instrumentalism.