Economic Skepticism



Download 163.46 Kb.
Page1/5
Date conversion03.05.2016
Size163.46 Kb.
  1   2   3   4   5
Chapter 2

Economic Skepticism

Economists and economics-minded observers are often skeptical about talk of “fairness” in trade. Often the concern is of a practical nature, that talk of fairness is open to abuse or even pernicious—a mere pretext for “protectionism” and a threat to liberal order of trade. Thus eminent trade economist Jagdish Bhagwati complains, in a similar connection, that the idea of “fair trade” is a “Pandora’s box” which has “grown out of hand.”1 We will return to such practical skepticism, and suggest that it does not call fairness into question once it is properly understood. Our focus in this chapter will be a different kind of skepticism which is less readily disposed of and which potentially strikes to the heart of our proposed conception of fairness itself. Answering this skepticism will unearth the social foundations of fairness, and so set the stage for the rest of this book.

We are proposing that fairness in the global economy be seen as (1) equity in the structure of a kind of international social practice, where (2) equity is assessed in light of that practice’s distributional consequences, within and across societies. But both of these claims can seem founded on bad economics, for reasons arising from nothing less than the standard case for free trade as understood since Adam Smith. This is for two main reasons.

First, the economic argument for free trade is said to be unilateral: each country has sufficient reason to free trade of its own accord, regardless of what other counties do. The standard reasoning is as follows. Trade augments national income because imports free up resources to be put in the service of more productive employments. Exports mainly pay for imports. In that case, the gains of trade can be reaped by freeing trade regardless of whether other countries do likewise; imports are still beneficial even if other countries protect their markets, and even if no general practice of freeing trade emerges. If in reality governments show reluctance to free trade, this merely shows that trade law and practice can be a useful remedial measure: it helps governments do what national prudence dictates in any case. As Paul Krugman forcefully puts the point, in criticizing trade negotiations for its misguided preoccupation with “mutuality” and “reciprocity”:

Anyone who has tried to make sense of international trade negotiations eventually realizes…they are a game scored according to mercantilist rules, in which an increase in exports—no matter how expensive to produce in terms of other opportunities forgone—is a victory, and an increase in imports—no matter how many resources it releases for other uses—is a defeat.2

On the contrary, Krugman explains, “If economists ruled the world, there would be no need for a World Trade Organization. (…) [G]lobal free trade would emerge spontaneously from the unrestricted pursuit of national interest.”3 And as it is with formal trade law it may be with mere practice as well. Trade will flower, provided only sufficient national prudence, whether or not mutuality in trade practice is ever established.

Second, even if international trade is a social practice, it is often said that our basic reasons to free trade make no essential reference to distributional concerns. The rationale for free trade lies in considerations of productive efficiency and aggregate national income. And if societies are concerned with the domestic “losers” from trade, it is often suggested, they can simply compensate them with some portion of the gains to national income. Distributional fairness provides no grounds for reluctance about more or less complete trade liberalization. As Nobel Laureate economist John Hicks famously counsels, “If measures for efficiency are to have a fair chance, it is extremely desirable that they should be freed from distributive complications as much as possible.”4

Upon closer examination, however, neither of these arguments undermines our proposed conception of trade as a shared, distriubtively consequential social practice. However strenuously the above points have been voiced by politically embattled economists, neither strictly follows from serous economic theory as applied to real world trade. And insofar as either or both do reflect many economists’ basic way of thinking about the global economy, this reflects a philosophical position which is readily open to doubt in light of certain basic facts of the human condition. To show this, it would be enough to show that our basic approach to fairness is fully consistent with economic theory. Our stronger claim will be that free trade argument depends on considerations of mutuality and distribution that are part and parcel of fairness issues of structural equity. We will find grounds in favor of our basic approach from within standard free trade argument itself.
TRADE AS A PRACTICE

We begin by further specifying our proposed conception of why fairness questions arise. A basic claim of this book, defended in this and subsequent chapters, is that the global economy is constituted and organized by an international market reliance practice, a social practice, created and sustained by nations, wherein nations each mutually rely on common markets.

To explain, let us say that a social practice exists when the following (sufficient) conditions hold:
(1) the behavior of two or more agents is coordinated over time (in contrast with intrapersonal practices involving the same agent at different times, e.g., brushing one’s teeth each night before bed);
(2) coordination among the agents is maintained by commonly known, mutually accepted behavioral expectations (in contrast, e.g., with purely accidental or unknown regularities in conduct);
(3) expectations are by and large set and fulfilled, despite potentially divergent interests, for the sake of a commonly understood and accepted end or aim (in contrast with coordination by pure force, backed by no publicly available coordinating rationale); and
(4) coordination is governable, if only by decentralized means, as applicable moral or fairness considerations require.
So, for example, social practices can be informal and of small scale (e.g., regular dining among friends; a household division of labor; practices for common resource use, in irrigation, fisheries, or forestry; the procedures of a deliberative body in an association or firm). Or they can be formalized and large enough to include strangers, as with the system of institutions that make up the modern state, or a practice of mutual aid on the open seas.

What, then, of international trade? Our initial answer is that all markets are “embedded” within nonmarket institutions or social practices, and that international trade is no exception.

Let us say that a market of some scope and kind is constitutively embedded within some nonmarket institutions when the very existence of a market of that that scope and kind requires that the nonmarket institutions be established and maintained. The standard Arrow-Debreu model of a competitive market,5 for example, presupposes a well-defined system of property and contract rights, as well as any security, judicial, and political institutions needed for such systems to be established and maintained. When such nonmarket institutions are wholly absent—when there are no understood property rights, no “yours” and “mine” to exchange—we do not have a “market failure” or an “inefficient market,” but rather fail to have a market at all; we have not a market, but something else, such as bartered exchange. 6 True laissez-faire is, in this sense, pure fiction. There could be no such thing as a purely self-organizing, self-sustaining, self-correcting market system of individual choices and exchanges (in the above standard sense) without some constitutively embedding institutions or other.7

The breakdown or absence of constitutively embedding institutions contrasts with mere market dysfunction, or failure of some functional standard, say, of efficient allocation. Here institutions may be required. When then there is “imperfect competition” or “market failure” (e.g., a negative externality or information asymmetry), well-functioning markets may require certain functionally embedding institutions (transparency laws or externality taxes).8 This can only be a modification within a larger social frame; functionally embedding institutions presuppose constitutively embedding institutions. For, again, when the market relation simply breaks down (e.g., because property rights disintegrate), we don’t have a market dysfunction to be cured. We don’t have a market at all, but rather bartered exchange, chaos, war, or something else.

As for international trade, most economists will perhaps hold that the global economy is constitutively embedded within the state system: distinct political authorities are in effect empowered to set their respective trade policies, each deciding to free, or not to free, cross-border exchange. Without this institutional background, we are not quite considering international trade, but something rather different.

That is not yet to say that the embedding institutions for international trade include shared international practice, as opposed to so many unilateral decisions to free trade. But trade comes to look more like a shared practice when we clarify the sense of “trade” that is our main concern. Our concern is with international trade of the sort that provides the centerpiece of the global economy as we know it. Many things often called “trade” are unsuited for that organizing role. Most obviously, the unilateral removal of trade barriers by a single country would not suffice, for the simple reason that other countries must open or have opened their borders as well. In at least that sense, trade depends on de facto coordination of government action; borders must be open at the same time. Nor will open borders and minimal exchanges qualify as “trade” in the relevant organizing sense. Trade flows must rise to the level at which the societies become substantially economically interdependent (even if only partially integrated), such that the pattern of exchanges affects the division of labor and overall structure of production. Moreover, even substantial trade flows will not, in themselves, be sufficient if they cannot be said to have a recognizably legitimate social purpose. Although socially unproductive “black markets” are in fact a large part of global trade flows, they exist largely because governments are unable to close the borders down. By contrast, as suggested in Chapter 1, international trade, in the relevant organizing sense, is assumed to have the presumptively legitimate social purpose of augmenting national wealth, by creating a shared, ever-refined division of labor, and ever-greater productive efficiency. Accordingly, the choice to maintain policies of open trade can be regarded as a kind of market reliance, in the sense that governments each choose, on behalf of their citizens, to rely on a common market as an organizing part of social life. In sum, we have “international trade,” in the basic, organizing sense of our concern, only when several countries become relatively integrated into common markets, because governments each rely on common markets, by allowing routine and sustained cross-border exchange, for the commonly understood purpose of mutual national economic gain.

This still not yet to say that trade constitutes a shared reliance practice; it still isn’t to say, that is, that market reliance is in any sense mutual or reciprocal, as opposed to mere de facto policy coordination. But this further claim becomes highly plausible when we add that no explicit reliance agreement is required, e.g., nothing like a formal arrangement of the form, “I’ll cut (or fail to raise) such-and-such tariffs if you cut yours.” It is enough for a shared practice that each government’s reliance on the common market is coordinated informally, by commonly known, mutually accepted expectations of market reliance, which are sustained on the basis of common purpose (the mutual augmentation of national wealth). While the mutual expectations may be implicit and unstated, if they were articulated, they might be put this way: “We both know that cutting such-and such tariffs is win-win.”

If that is right, then our four conditions for a social practice are satisfied as follows. As for condition (1), the multiple agents in question are countries as represented by their respective governments. Government trade policies are coordinated over time, in the sense that each government concurrently maintains policies needed for a common market to exist. These include forbearance in the use of “border measures” such as tariffs and quotas, but also in “behind the border measures” such as subsidies, internal taxes, or preferential rules which undercut the effect of opened borders. As for condition (2), such policy coordination is not a mere de facto behavioral regularity, but undertaken and sustained on the basis of commonly known, mutually accepted market reliance expectations. Beyond minimum expectations of the trade policies needed for a common market to exist, participation includes compliance with expectations regarding the form that market reliance takes, as specified by formal treaty rules (e.g. a rule forbidding quotas, binding tariffs, or limiting subsidies) as well as informal understandings about how each country is to strike the balance between market and state (the “embedded liberalism compromise,” or current WTO practice). As for condition (3), such market reliance expectations are set, and by and large fulfilled, despite potentially divergent interests in using trade policy for “beggar-thy-neighbor” gains (of which more below). Expectations are set and fulfilled according to the commonly known and mutually accepted aim first specified by classical trade theory: the mutual augmentation of national (average or aggregate) income. As for condition (4), market reliance expectations are adjusted, and the larger practice so governed, in several decentralized ways: by formal multilateral negotiations within the WTO; informal multilateral negotiations (the OECD, the G20); formal bi-lateral or regional negotiations (regional or bi-lateral trade arrangements such as the EU or NAFTA); and informal bi-lateral negotiations (diplomatic or central bank suasion). Along with such “official” forms of governance, we may add public pressure and advocacy by non-governmental organizations and special interest groups (e.g., aid agencies, unions, multinational firms).

In other words, even perfectly free international trade is, and can only be, constitutively embedded within an international practice of mutual market reliance, much in the way a domestic capitalist economy is, and can only be, constitutively embedded within domestic institutions. But for a practice wherein countries each rely, mutually, on common markets, international commerce would not and could not go on in anything like the extensive way it does (residual “black markets” aside). Thus complete “free trade” is but one version of market reliance—indeed, a version only somewhat more “free” than the relatively liberal market reliance practice that governments now sustain.9

In this case, we have a proper subject of fairness. So long as there is a practice of mutual market reliance, it is appropriate to ask the question of structural equity: How must the market reliance practice be set up if the way it distributes its advantages and disadvantages is to be reasonably acceptable to each country and class it affects? Questions of “fairness in trade” are, in that sense, both apt and potentially of great practical significance for how functionally embedding, nonmarket institutions are arranged.


TRADE AS AN EXCEPTIONAL CASE

Of course, the foregoing is by no means the standard picture in professional economics. Most economists would admit that domestic capitalist markets require various nonmarket institutions, and perhaps that international trade depends on the general state system and its associated rights of property, sovereignty, and so on. But at least on one standard picture, international markets otherwise present an exceptional, perhaps derivative case: they do not essentially depend upon international economic law or practice. As suggested above, the standard conception of why gains arise from trade suggests that no such practice is required. Again, the gains arise from imports which free resources to do more productive things, and exports mainly pay for imports. Or to put much the same idea in more sophisticated terms, neo-classical models of trade (e.g., the Heckster-Ohlin-Samuleson model) show net national income gains without cooperative law or practice of any kind. The suggestion, then, is that free trade really would emerge from mere national prudence, if countries could simply get the economics straight and pull the levers of power accordingly (being free from confusion, special interest pressures, faintness of heart, and so on).10 Given prudence, reality would sufficiently approximate the models.

To be sure, this picture is oversimplified even from the point of economic theory (though in much the way economists oversimplify it in public debate). It nevertheless presents a significant challenge to our basic thesis that the global economy offers a significant site of fairness assessment in its organizing international social practice, a practice of which the current the multilateral system of trade is one possible expression.

To see the force of the challenge, consider one line of reply which accepts the suggested basic picture. The reply is as follows. So long as a trading system does already exist, we do have a market reliance practice that can be assessed as having an equitable or inequitable structure. Its origins in imprudence are immaterial. The case is then much like discriminatory law enforcement: if one racial group is much more likely than others to be prosecuted under established laws against murder, with no actual group difference in murder rates, the charge of structural inequity will be apt. And the charge will be apt even if such laws are rooted in the imprudence of people—that is, even if, in a world of perfect prudence, no one would commit murder given the risks of personal injury, reputation costs, harm to one’s soul, and so on. So long as laws against murder do exist, discriminatory enforcement is an important kind of structural inequity, which a society would have decisive reasons to reform. Likewise, even if the WTO system is borne of national imprudence, this does not preclude structural equity assessment of how it distributes any advantages and disadvantages that it in fact creates. Structural inequities may even be of the first moral importance.

This is, however, a pyrrhic victory. Even if structural equity concerns would not be wholly out of place, their scope would remain sharply limited so long as the standard economic picture remains unchallenged. On the standard view, international economic cooperation might help nations help themselves, but it would otherwise have a limited role in explaining the outcomes that actually ensue from trade. The major economic outcomes (for unemployment, or national income) are not attributable to international cooperation per se. The gains of trade are less as the result of an international joint enterprise than something akin to climate management: each government adapts to its local weather, for good or for ill, but the goods or bads received (barring cross-border pollution, greenhouse gas emissions, etc.) do not ultimately reflect its participation in a common practice with essentially social fruit.

One of our main theses, however, is that the gains of trade are generally the fruit of international cooperation. We therefore mount a stronger line of reply: a mutual market reliance practice is an essential enabling condition for most any trade relationship that could qualify as the centerpiece of the global economy as we know it. The practice is essential, not because of national imprudence, but because of unavoidable problems of uncertainty and risk that arise in any social action among distinct agents. In short, economic reality cannot approximate neo-classical models in the absence of a constitutively embedding, shared market reliance practice. The gains of trade, in the most basic sense, are neither pure self-help nor akin to local climate management policy, but by their very nature socially created through international cooperation. They are therefore within the proper ambit of international structural equity.

To see why this is so, observe the grain of truth in the standard picture: according to economic models, free trade is to each country’s unilateral advantage, whatever other countries do, when (i) governments care only about maximizing national income, and so are indifferent to domestic distribution (at least as far as trade policy is concerned), and (ii) no country is able to impose an externality upon the others, e.g., by altering the “terms of trade” (i.e., the buying power of exports for imports). Now, even if the model applies in certain special situations (e.g. a “small” country integrating into a larger economy, with an ample tax base and solid compensation schemes), the central question is whether the model is representative of the global economy as we know it. But it isn’t especially representative, for general, fundamental reasons.

For one, almost every country, always and everywhere, cares about domestic distribution (if only because of special interest pressure). And this is as it should be. As we will see below, standard free trade argument depends on sensitivity to domestic distribution (e.g., on whether “losers” will in fact be compensated).

Moreover, “mutuality” and “reciprocity” in trade are, in a very basic sense, unavoidable in any trade relationship. According to important recent work in economic theory, terms-of-trade (or other) externalities are in themselves substantial enough to justify “mutuality” or “reciprocity” in trade agreements.11 Indeed, much of the existing multilateral system can be explained in game-theoretic terms, as a way of escaping what Bagwell and Staiger call a “terms-of-trade prisoner’s dilemma.”12

It is important, however, that such work offers only a partial defense of “mutuality” in the sense denied by the above picture. If nothing else is said, the initial, standard picture can again simply take a more qualified form: even if formal agreements, such as the WTO, can be required to optimize trade gains, it may be said, that is not to say that informal practice is in any sense fundamentally required. We will see, however, that recent game-theoretic accounts do not fully develop their own fundamental insight, which is not essentially about formal arrangements. It is common practice, whether formal or informal, that is essential: virtually any lasting trade relation depends on the establishment of a shared social practice of trade marked by (perhaps implicit) mutuality. A practice of trade is essential for the resolution of basic problems of assurance which arise from conditions the human condition itself.


THE STANDARD CASE FOR FREE TRADE

We present our argument for this momentarily. In order for it to be at all convincing, it must be clear that we are taking the standard economic case for free trade very seriously. To that end, as well as to review the case for non-specialist readers, we pause to reconstruct the standard case in detail.13

Why should countries trade? Or more specifically, why should the governments of different countries remove trade barriers—quotas, tariffs, duties, subsidies, controls, and so forth—and so allow transactional flows in goods, services, or capital to cross their borders? They should do so, according to standard economic theory, because each country thereby augments its national income, taken in the aggregate. One of Adam Smith’s signal advances over mercantilist thought (which called, e.g., for export-promotion and import-minimization to protect jobs) was the benchmark of national income.14 Economists have since sounded a common refrain: imports free productive resources to do other things. Exports mainly pay for imports. The losses that imports create are more than outweighed, in the aggregate, by the resulting productivity gains.15

  1   2   3   4   5


The database is protected by copyright ©essaydocs.org 2016
send message

    Main page