|
Article Excerpt
INTRODUCTION
Administrative issues do not usually make for exciting economics. They usually fall into the forgotten footnotes of scholarly articles that are primarily concerned with grander visions of economic mechanisms. A discussion of the possible impact of developments in information technology for the implementation of government policy might seem to fall into that category. However, it is also possible to identify key innovations in recent history that have become not just background items in a more interesting economic story, but, indeed, the story itself.
So, within the context of tax policy, is the "smart card" a convenience in doing something a little more effectively than we already do it, or is it potentially a dramatic new opportunity? In this paper we examine some of the economic arguments that bear on these questions, drawing on the standard analysis of direct and indirect taxation. The principal focus is on some fundamental informational constraints in the implementation of optimal tax systems--do smart cards hold the promise of alleviating any of these constraints? However, although the argument is simply from a conventional economics perspective, I do not wish to underestimate the potential contribution to the discussion from related fields such as accountancy and law.
CARDS: SMART AND NOT-SO-SMART
At the outset it is as well to try to develop a taxonomy of the kinds of cards that we are talking about. What should we assume that they can do?
Let us say a word about non-smart cards first. In their certification role, dumb cards may have considerable promise and may assist in overcoming a class of information problems that we shall review in the fifth section. However we should acknowledge that their contribution could be achieved in other ways--paper documents such as identity cards or passports, biometric records, tattoos. We may leave open a number of questions that could be relevant to economists, lawyers and administrators, such as the visibility of the information and the degree of compulsion in use of the card. Effective certification clearly could make a major contribution in terms of tax administration and possibly in terms of enforcement but, in one sense, it is nothing new.
If we want to discuss smart cards, then what do we mean by "smart" and what more might smart cards do? The focus presumably ought to be on "smartness" in the encoding of information that is potentially useful in tax design. Let us distinguish three ways in which the encoding can be said to be smart:
* Quantity of information. In a snapshot sense it can appear to offer considerable precision. It is clearly a lot more convenient than having something tattooed on your forehead. But in essence it is not really doing anything different from a dumb card, just doing it more intensively and comprehensively.
* Automatic updating. A natural extension to the basic idea of the first smart-card role. It can work in either or both of two principal ways. The first of these takes account of the potential changeability of exogenous information: what makes the smart card smart is its adaptation to new circumstances. The second way it can update itself is in recording the bearer's decisions and choices. Here the card acts as a small computer database. Again the role is foreshadowed in old-fashioned paper records: as an example think of endorsements to a driving license triggered by a traffic violation. (1)
* Information linking. The third smart role for a card is as an electronic master key. The card itself does not carry the database but is a conduit of approach to official confidential databases. We will return to this issue in the fifth section.
I suggest that the key distinction is not whether cards are dumb or smart, but whether cards are static or dynamic. In the static case, the role of the card is summarized in an electronic statement: "this certifies my characteristics;" it is just the electronic tattoo and it may be convenient for both individuals and officials. However, dynamic cards could perhaps achieve something more subtle. Their mission statement is something like: "this certifies the history of my transactions" or "this certifies the history of my characteristics," or perhaps both statements combined.
As we will see, all three versions of smartness are of potential interest in questions of tax design. Some of them appear to present a considerable challenge.
QUESTIONS ABOUT TAXATION
In order to examine the potential of smart cards, let us briefly review how taxes work.
Taxation: Necessity and Objectives
Let us begin with two basic questions, the answers to which may seem rather obvious.
First, why do we have taxes? There are, broadly speaking, three principal reasons advanced in the literature.
1. There are socially desirable goods that cannot be charged for and the funds to provide these have to come from somewhere. The classic example of this is the provision of national defense.
2. There are cases where price manipulation is in the public interest, to correct market "distortions." (2) Examples of this are the use of taxes to achieve environmental objectives or to offset other effects of consumption and production externalities.
3. It may be considered appropriate to use fiscal measures to implement other forms of social engineering; this often involves manipulating the distribution of income or of spending power.
Second, what economic principles should guide the design of taxes? The conventional list of criteria consists of allocative efficiency, distributional equity, and effective administration.
These two questions are relevant for understanding the potential for smart cards. Cards could, in principle, have a role in all three taxation roles cited in the response to the first question. However, while it is clear that technological improvements embodied in cards could potentially improve the effectiveness of tax administration, the case remains to be made as to whether there is a similar advantage with respect to the efficiency and equity objectives.
Tax Types and Information
Now a less obvious question, also relevant for cards: why is there a separation between direct and indirect taxation and what is the nature of the separation?
This is not just a matter of labeling nor of administrative convenience. (3) A more fundamental reason lies in the structure of information associated with each broad type of tax and the scope for the shaping economic incentives. Put simply: what is a government agency allowed to know or what should a tax-modeler assume it is allowed to know? In the case of direct taxes, the appropriate assumption is that the tax agency is well informed about the attributes and circumstances of individual persons, including their income; but there is severely limited detail as to the disposition of personal income across alternative expenditure categories. In the case of indirect taxes, the agency is well informed about broad classes of transactions, by commodity group, by industry sector, perhaps by firm; but detail on purchases by individuals may be restricted.
This discussion leads to a crucial further question: are both types of tax necessary? As we will see below (in the second subsection of the fourth section), if we follow the standard economic theory of taxation, there is a strikingly direct answer to this question. One issue that we need to examine is whether the presence of the smart card is likely to modify the answer to the question.
Taxes in Practice
Before we develop this analysis using a proper model, let us also review the way taxes work in practice. Let us take an illustrative example from the UK, focusing on the broad categories of direct and indirect taxes. The UK case is not particularly special but...
|
