Revue Européenne du Droit
The Right Scope of Global Governance and Democracy Enhancement
Issue #2


Issue #2


Dani Rodrik

21x29,7cm - 186 pages Issue #2, Spring 2021 24€

Greater interdependence is often taken to require more global governance, but the logic requires scrutiny. Cross-border spillovers do not always call for international rules 1 . The canonical cases for global governance are based on two sets of circumstances: global commons and “beggar-thy-neighbor” (BTN) policies. The world economy is not a global commons (outside of climate change), and much of our current discussions deal with policies that are not true BTNs. Some of these are beggar-thyself policies; others may produce domestic benefits, addressing real market distortions or legitimate social objectives. The case for global governance in such policies, I will argue, is very weak, and possibly outweighed by the risk that global oversight or regulation would backfire. While these policy domains are certainly rife with failures, such failures arise not from weaknesses of global governance, but from failures of national governance and cannot be fixed through international agreements or multilateral cooperation. I advocate a mode of global governance that I call “democracy-enhancing global governance,” to be distinguished from “globalization-enhancing global governance” 2 .


In a world economy that has become highly integrated, problems always seem to require more international cooperation and better global governance. The populist backlash, as well as ex-U.S. President Donald Trump’s trade policies, have added fuel to calls made by economists, technocrats, and commentators for more internationalism. “[V]irtually every problem destabilizing the world in this plastic moment is global in nature and can be confronted only with a coalition that is global . . .”, wrote the New York Times columnist Thomas Friedman recently 3 . Or as Nemat Shafik, then the Deputy Managing Director of the International Monetary Fund, put it in 2013, “What happens anywhere affects everybody—and increasingly so.

So it is pretty clear that the world needs more, not less, international coordination and cooperation” 4 . When the European economics network solicited advice from leading economists on how address the frailties of the global financial system in the wake of the 2008 crisis, the proposed solutions often took the form of tighter international rules administered by some kind of technocracy: an international bankruptcy court, a world financial organization, an international bank charter, or an international lender of last resort 5 . Nationalism may seem ascendant in politics, but global governance reigns in economics.

It is tempting to think that greater interdependence requires more global governance, but the logic requires scrutiny. On the one hand, interdependence blurs the distinction between what is domestic and international. On the other hand, there remains considerable institutional diversity among nations, rooted in different historical, cultural, or development trajectories. This diversity reflects itself both in different preferences (“objective functions”) and in different perceptions of how the world works. These in turn make it difficult for countries to agree on common policies or rules. Today’s U.S.–China trade conflict is the paradigmatic example of the tensions that arise in the absence of a satisfactory solution to this dilemma 6 . When should global rules override national differences and impose common solutions?

Consider the following policies: educational policies; highway speed limits; gasoline taxes; agricultural subsidies; import tariffs on cars; and tax havens.

In an economically interdependent world, each one of these policies produces spillovers—or cross-border externalities—to other nations. The last three policies are typically considered international, and subject to global governance. The first three are normally considered “domestic” policies, but they too have global implications. Educational policies can shape a country’s comparative advantage and will thereby influence its (and other countries’) terms of trade. Highway speed limits and gasoline taxes affect the domestic demand for oil and therefore the price of oil on world markets. The presence of cross-border spillovers does not seem like a sufficient condition for global governance.

In fact, it is not at all clear how the dividing line that is conventionally drawn between the two sets of policies is actually drawn. Should we focus on the magnitude of cross-border spillovers? This is an empirical matter requiring case-by-case analysis. For example, taxes on gasoline in the United States and Europe likely have far greater impact on world markets than auto tariffs in small or medium-sized countries. Should we ask instead whether there is harm to other nations? But export subsidies on farm products are beneficial on net to the rest of the world since they deteriorate the subsidizing country’s external terms of trade and improve the terms of trade of the rest of the world. Perhaps we should focus on the stated objective of policy—domestic versus international? Yet educational investments are often justified on the grounds of increasing a country’s global competitiveness, making them as international in this sense as trade policies. None of these criteria does a good job of explaining why the first set of policies is “domestic” and the second is “international.” The muddle of global governance is that many policies have become “internationalized” through happenstance or the operation of political lobbies, rather than on account of principled distinctions.

The canonical cases for global governance are based on two set of circumstances 7 . The first occurs when there is global public good (GPG). The classic case is carbon control policies in the presence of climate change. The second is represented by “beggar-thy-neighbor” (BTN) policies. A BTN policy is one that produces an income transfer to the home economy from the rest of the world while producing global inefficiency as a by-product. Exploiting monopoly power in a rare metal on global markets by restricting sales abroad would constitute an example. Both of these circumstances provide impeccable arguments for global governance establishing and enforcing guidelines on what countries can do on their own. However, their relevance to the burning policy issues of the day is much more limited than is commonly realized. As I will show below, the world economy is not a global commons, and virtually no economic policy has the nature of a global public good (or bad). And while there are some important BTN policies, much of our current discussions deal with policies that are not true BTNs. Subsidies, industrial policies, employment-protecting tariffs, non-tariff measures that target health or social concerns, poor financial regulations, inappropriate (excessively austere) fiscal policies, or national Internet walls are neither GPGs nor BTNs. Some of these are beggar-thyself policies; others may produce domestic benefits, addressing real market distortions or legitimate social objectives. The case for global governance in such policies, I will argue, is very weak, and possibly outweighed by the risk that global oversight or regulation would backfire.

None of this is to suggest that we live in a Panglossian world where all policy is for the best. The policy domains I have just lifted are certainly rife with failures. When large countries make serious policy mistakes—as in the case of the United States with lax financial regulation in the run-up to the global financial crisis of 2007–2008— other countries pay a price as well. My argument is that such failures arise not from weaknesses of global governance, but from failures of national governance. These failures cannot be fixed through international agreements or multilateral cooperation. External constraints may in fact aggravate domestic failures of governance, insofar as they empower particular distributional coalitions at the expense of the broad publics. At the end of this paper, I advocate a mode of global governance that I call “democracy-enhancing global governance.” Unlike “globalization-enhancing global governance,” democracy-enhancing global governance would leave most policy domains to national regulation, with global oversight restricted to procedural safeguards—such as transparency, accountability, the use of scientific/economic evidence—intended to reinforce democratic deliberation.

The Analytics of Economic Interdependence:

The Case for Global Governance

It helps to have an explicit framework to discuss the issues that arise in the presence of spillovers and to distinguish among different kinds of problems 8 . By thinking of these issues in terms of utility functions of the relevant countries, it cannot be overlooked that well-being at home and abroad is affected not only by a country’s own policies, but also by (the sum total of) policies of foreign countries. Spillovers could be negative or positive, depending on the policy in question. When countries act independently, maximizing their own utility and disregarding the effects of their choices on other countries, we have the standard result that the resulting (Nash) equilibrium will be inefficient. Policies with negative spillovers would be oversupplied, and policies with positive spillovers will be undersupplied. Pigovian taxes and subsidies that enable the internalization of these cross-border externalities are obviously impractical in this context.

There are two benchmark cases where all countries could be made better off through global rules that discipline home and foreign policies. I take them up in turn.  

Global Public Goods (GPGs)

Suppose that in addition to their direct, domestic effects, home and foreign policies jointly contribute to provide a global benefit (or damage), which is non-rival and from which individual countries cannot be excluded. Suppose also that, for an identical set of policies, the contribution of each country to the global benefit (or damage) is proportional to its size.

In this case, it can be shown that when countries are small, they overlook the effect of their policies on the global benefit (or damage). Since these effects are systematically discounted, a global public good will be underprovided and a global public bad will be overprovided. The best-known example of this is greenhouse gas (GHG) emission, a global public bad in view of climate change. Policy here consists of controls on GHG. Since such controls are costly at the domestic level while providing benefits only in terms of global benefits, countries will have the incentive to minimize these controls. A global agreement that capped domestic GHG emission levels would leave all countries better off, assuming countries are sufficiently similar, or the caps are appropriately calibrated to individual country circumstances.

When commentators describe the world economy as a “global commons” or free trade as a global public good, they have a similar argument in mind. But this is a misleading analogy. Economic policies that are beneficial to the world economy also tend to be beneficial for the home economy; they are primarily private, rather than public goods.

First, consider trade policy. It could well be that open trade policies contribute to a global public good: the benefits from trade may increase with the number of countries that practice free trade. But the relevant question is whether a country that disregards this external benefit would have the incentive to pursue globally suboptimal trade policies. For a small country, the answer is no. Free trade is the optimal policy for domestic reasons, regardless of other countries’ policies. In other words, domestic utility is maximized when tariffs and non-tariff barriers are set to zero. (I will take up the large country optimal tariff case later.) This is very different from the GHG case where policies are costly at the domestic level, and the home country wants to set GHG controls at their lower limit. Countries trade not to confer benefits on their partners, but to reap the domestic gains from trade. And when they forsake those gains from trade, the problem is not with lack of global governance; the much larger failure lies at home, with domestic governance.

Much the same logic applies in many other policy domains where good economic policy is its own reward. Consider financial markets. Prudential financial regulation ensures that financial intermediaries do not take on too much risk and financial instability is kept in check. When financial centers pursue appropriate policies, they enhance financial stability and soundness for the global economy as a whole. But these centers have all the incentive in the world to adopt such policies since they will be the first to bear the costs of financial crises. The 2008 global financial crisis may have been due to lax financial regulations in the United States. But these policy mistakes did not originate from the U.S. government’s lack of concern for the global economy. Rather, they were the result of a series of misjudgments with respect to the domestic consequences of financial liberalization. U.S. regulators did not require greater cosmopolitanism; they needed a better sense of the national interest.

Similar arguments can be made for fiscal policy, tax policy, and regulation in general. Excessive austerity can be damaging to the world economy, but the costs are borne first and foremost at home. Inappropriate tax policies or poorly designed regulations hurt the home economy in much greater measure than they affect other nations. In all these areas, policies that sustain a healthy global economy are—or should be, with appropriate domestic governance—in the national interests of each country. The extent to which global governance can help fix domestic governance problems is a different question, to which I will turn later. The point for now is that most standard economic policies cannot be considered to be GPGs.

Beggar-Thy-Neighbor Policies (BTNs)

Beggar-Thy-Neighbor policies provide benefits at home only to the extent that they harm other countries. Additionally, they generate global inefficiency, a deadweight loss. A well-known instance in trade policy is the so-called optimum tariff, whereby a large country can manipulate its terms of trade by restricting its imports (or exports). Since other nations face similar incentives, in the end all (or most) countries end up worse off by engaging in destructive trade practices 9 . This type of problem represents the second canonical case for global governance. Distinguishing between BTN and non-BTN policies can be useful in practice to ascertain the limits of desirable international rules. For example, my colleagues and I have proposed an approach to U.S.-China economic relations that centers on drawing a bright red line around BTNs 10 .

Take a two-country example. The utility functions of the two countries can be represented as the sum of two components, a regular part that depends only on own policies and a second part that captures the pure transfer component of the policy. Note that in this case the transfer component is zero-sum: whatever home gains from its policy comes at the expense of losses to foreign, and vice versa. Assume for simplicity that the two countries are identical. Then in equilibrium, neither country is able to extract transfers from the other. But in attempting to beggar each other, they are both driven to take inefficiently high levels of actions. This is exactly what happens in the optimum tariff case.

Another example would be the use of mercantilist currency policies. Assume there is generalized unemployment, and both countries would benefit by running a trade surplus. Each country tries to undervalue its currency or follow other policies to improve its trade balance. But one country’s trade surplus is the other country’s trade deficit. In the end, such efforts offset each other. Neither country ends up with higher employment, but both suffer the incidental costs of mercantilist policies.

Since BTNs are negative-sum policies, there is a strong presumption that they should be restrained using global rules. Note that in the two examples I have used above, it is also the case that both countries are better off when their policy autonomy is restricted (by placing ceilings on possible unilateral actions) compared to when they have full autonomy. Subject to the usual caveats about commitment, these are the relatively easy cases for global governance 11 .

But there are other cases of BTN policies where one or more of the countries would be worse off in the cooperative equilibrium. (Side payments from the beneficiaries to the losing country would rule out such a possibility, but these are difficult to implement in a global context.) In the optimum tariff game I discussed, in the presence of asymmetry it is possible for one of the countries to prefer the Nash equilibrium to the cooperative equilibrium: a larger country gains more from manipulating its terms of trade than a smaller one and has more to lose from international disciplines. The example of a global cartel, mentioned in the introduction, is another example.

Suppose a number of exporters of a key commodity have cartelized and are facing a large number of small importers. A cooperative equilibrium that prevented them from exercising monopoly power would definitely leave the cartel members poorer. In this instance, there is no incentive for these countries to join any global governance scheme.

A second, similar example is that of pure tax havens. A pure tax haven is a jurisdiction that applies a very low corporate tax rate for the sole purpose of enabling international corporations to engage in tax evasion. This represents a BTN policy because it undermines the tax base of countries and shifts the global tax burden towards labor, a poorer group, without stimulating physical investment. Pure tax havens shift paper profits to low-tax jurisdictions, not physical capital 12 . In this case too, global governance that prevented tax competition would leave some countries, namely the pure tax havens, worse off. They would be deprived of the revenues they generate by attracting a very large base of paper profits at very low tax rates. An analogous case can be made for personal income or wealth tax havens. A global registry that would identify ultimate owners of bank accounts in all financial jurisdictions would assist tax administration and collection and benefit most countries of the world, but the tax havens would lose out.

Whether they make all countries better off or not, the demands that BTN policies make on global governance are rather limited. That is because relatively few policies fall under this rubric. In fact, I have mentioned here all the straightforward examples of BTN economic policies I can think of: optimum tariffs, international monopolies or cartels, trade-balance mercantilism, and pure tax havens (for corporate and personal income) 13 . Perhaps U.S.-China competition in digital technologies opens new areas of BTNs, but the vast majority of economic policies that are contentious and come under international scrutiny are not BTNs, even though they are frequently presented as such—just as the global economy is often misleadingly viewed as a global commons.

The Weak Case for Global Governance: Policies that are Neither GPGs Nor BTNs

Consider the following two policies: R&D subsidies in a country that imports technology-intensive goods; and an import ban on goods produced with slave labor. Both policies create negative cross-border spillovers. The first improves technological capabilities in the home economy and can be expected to have an adverse terms-of-trade impact on the rest of the world. This is because as the country becomes better at producing technologically advanced goods, its demand for imports of such goods fall. The second policy has a direct adverse economic impact on exporters of slave-made goods. In both cases, current practice is that such policies are not regulated internationally. Countries are free to do what they please in both domains. My guess is that this conforms to the intuition of most analysts with respect to what constitutes an appropriate dividing line between domestic and international spheres of regulation. I will argue in this section that a large number of policies that global policy makers do try to bring under global governance are precisely of the same nature as one or the other of the two examples just mentioned. In particular, they have the following characteristics: (a) they either do not create global inefficiency; (b) or when they do, it is the domestic economy that bears the direct economic costs.

Technology subsidies are in category (a), assuming there are knowledge spillovers (even if these spillovers are purely domestic). The reason global governance is not believed to be appropriate in this instance is presumably that there is an economic justification for the policy in question, and the presence of cross-border spillovers is not grounds on its own for limiting what each nation can do independently. The import ban is in category (b). The reason for allowing a country wide latitude in this case is different: an import ban might be economically inefficient, but it is the home country that pays the economic price for it first and foremost. Effectively, the home country trades off the economic cost against the value of upholding a moral standard against slavery. It does not seem fitting for an international organization or a global governance regime to second-guess the appropriateness of this tradeoff.

Yet many other policies that are routinely internationalized are no different. I will examine them below.

Non-BTN Policies: Enrich-Thy-Neighbor Policies

Enrich-Thy-Neighbor policies produce positive aggregate effects on the rest of the world and are yet contentious globally. This seems paradoxical, and it is. There is one significant category of policies that fit this description: subsidies on exportables. Whether they be on agricultural products or manufactured goods, export subsidies are considered to be a no-no internationally. This is puzzling since export subsidies are an economic “gift” to the rest of the world. True, some foreign countries may lose, but this does not alter the fact that the aggregate effect on the rest of the world is positive 14 .

Consider a world with two foreign countries. Assume that an increase in certain policies of the home country has asymmetric effects abroad. One foreign country benefits, while the other country loses. When the policy at hand is an export subsidy, we know that the sum of these two effects of opposite sign has to be positive. That is because the export subsidy deteriorated the home country’s terms of trade, and therefore improves the terms of trade of the rest of the world in aggregate (the two foreign countries taken together). The asymmetric effects in turn would be due to the pattern of comparative advantage across countries. One foreign country may have a comparative advantage structure similar to the home country, so that its terms of trade and those of the home country move together. So agricultural export subsidies, for example, will make net exporters of agricultural products in the rest of the world worse off, while making net importers better off.

There are three arguments for why export subsidies should nevertheless be globally disciplined, none of which is very compelling. First, it is the case that some foreign countries lose. Members of the Cairns group of large agricultural exporters have made their case loudly and successfully within the GATT/WTO regime in the case of agricultural subsidies. But this is a curious argument insofar as there are multitudes of policies that are left under national prerogative, but likewise produce asymmetric effects abroad. These include policies that are roundly applauded by economists and technocrats as appropriate policies. Consider unilateral import liberalization as a blatant example. When a large country unilaterally reduces its import barriers, it normally incurs a terms-of-trade loss. More to the point, foreign countries that share this country’s comparative advantage pattern also experience a terms-of-trade loss. (When I increase my imports of textiles and autos, driving their relative prices up on world markets, all other net importers of textiles and autos suffer too.) And in their case, there is no compensating increase in gains from trade. So, some foreign countries are definitely left worse off. To my knowledge, this has never been used as an argument for placing global limits on countries’ ability to unilaterally liberalize their trade regimes.

The second argument is that subsidies, unlike unilateral import liberalization, are globally inefficient. This justification for global governance has to do with the economic desirability of subsidies in general, and not with the incidence of their external effects. The trouble here is that it is difficult to take such a categorical stance against the use of subsidies. There may be genuine learning externalities associated with exporting, which the subsidizing country aims to reap. Or there may be social or political objectives that are equally justifiable on broader grounds, even if not strictly economically. Just as the moral stance reflected in the import ban on slave-produced goods cannot be second-guessed by other countries, it may not be appropriate for foreign countries to question whether a particular social objective is valid or best addressed through subsidies. I will scrutinize these issues at greater length in the next subsection. What can be said unambiguously is that when the subsidies do not serve a real economic purpose, unlike BTNs, their most immediate costs are shouldered by domestic taxpayers and consumers.

The third argument is that subsidies (and similar policies) are “unfair” because they undermine level playing fields in global trade. This argument is based on the view that all nations should compete on an even basis 15 . But what constitutes a level playing field, like fairness, is very much in the eye of the beholder. For example, developing nations have long made the argument, not entirely unreasonably, that subsidies (like lax patent rules) serve to compensate for the disadvantages of backwardness, and in any case are practices that advanced nations themselves pursued when they were poorer. In other words, they make trade fairer rather than less so. Of course, if global agreement can be reached on what is “fair,” it makes sense to pursue common standards. But often such common ground will be lacking. In those cases, it would be inappropriate to seek global disciplines.

This is not to say that “unfair trade” is an empty and useless concept. When nations face trade transactions that they think undermine domestic moral codes or norms of fairness, they should be free to regulate them accordingly 16 . The difference is between living by one’s own moral standards and imposing them on others 17 .

Non-BTN spillovers: Policies with Ambiguous Domestic Efficiency Implications

We finally consider policies that produce adverse spillovers to other countries but are used for domestic reasons rather than for BTN purposes. These domestic reasons might be economic or non-economic, well-grounded or not. There is a very wide variety of such policies that are either already regulated internationally or frequently come under international scrutiny. Here is a partial listing:

▪  “weak” intellectual property rights protections,

▪  industrial policies that do not involve export subsidies, such as domestic subsidies, local content requirements, “trade-related investment measures,” etc.,

▪  bans on GMOs, hormone-fed beef, and other similar “health” measures,

▪  “excessive” fiscal austerity,

▪  “lax” financial regulation,

▪  import protection to prop up employment in certain industries or regions,

▪  “very low” levels of corporate taxation (i.e., as in Ireland, where there may be effects on domestic capital formation, not pure tax havens such as the Cayman Islands);

▪  data localization, local-cloud policies, and other Internet-nationalizing policies.

The domestic economic effects of all these policies ex ante are either negative or ambiguous. And they typically generate negative spillovers for other countries. All of this may suggest a rationale for global governance in these areas. The difficulty is that, as in the export subsidy case, there are strong countervailing arguments that cannot be dismissed. First, while policy failures are obvious ex post, they may not be so clear ex ante. To take a prominent example, there is now widespread agreement that financial regulators in the United States did a poor job of reining in risk-taking in mortgage lending and in the shadow banking sector prior to 2007. But views on the appropriateness of prevailing regulatory practices diverged significantly at the time. It is not at all clear that greater international coordination on financial regulation would have produced better outcomes. In fact, given the influence the U.S. financial sector exerted on the determination of Basel rules, the opposite was just as likely.

Second, there may indeed be market failures of distortions at home that justify the use of such policies, as second-best remedies even if not first-best ones. It is not obvious that trade negotiators or international bureaucrats are better placed than domestic legislatures and policy makers to make the right call in complex cases. Third, there may be overwhelming non-economic considerations—social, environmental, health, national security or moral—that trump economic costs and benefits. Once again, the relevant trade-offs are better evaluated at the national level, within pre-existing democratic decision-making bodies, than via delegation to international agencies.

The primary argument for global governance in these cases, one that economists are especially fond of making, is that global rules can prevent countries from using “beggar thyself” policies by correcting domestic political failures. A more sophisticated political-science version is provided by Keohane, Macedo and Moravcsik 18 ; it is a version of the standard delegation argument. Essentially, it views external constraints acting as a counterweight to special interests or rent-seeking lobbies. Trade agreements, for example, allow governments to say “no” to their protectionist lobbies at home by adhering to the following logic: “we would love to raise tariffs on this product, but WTO rules do not allow us to do so.”

There are three counterarguments. First, non-standard, heterodox policies can have economic justification in second-best contexts 19 . Global rules or bureaucracies cannot reliably distinguish between “beggar thyself” and economically desirable policies. This is especially true in policy domains that require significant local knowledge, as with industrial policies or financial regulations. Second, even when there is a strong presumption that countries are engaged in “beggar thyself” policies, democracies should be allowed to make their own “mistakes.” For example, the European Union may be deluded in banning GMOs or hormone-fed beef, but allowing supranational bodies to pass judgment in such matters undermines both democracy and the legitimacy of global governance arrangements. Third, and perhaps most importantly, there is no presumption that global governance institutions are more immune to capture by special interests than domestic policymaking. Indeed, large corporations, international banks, and Big Pharma have exercised disproportionate influence on global economic governance. It would be naïve to presume that they have prioritized the public interest over their particular interest in shaping global agreements in line with their needs.

I have developed the last point in a previous publication in the context of trade agreements 20 . The conventional view of trade agreements is that they offer welcome relief against protectionist interests at home, that is, inefficient import-competing firms and labor unions. When trade agreements were largely about import tariffs and quotas—that is before the 1980s—this made a lot of sense. Multilateral trade negotiations were about lowering these barriers, which meant going against what protectionist interests at home wanted. But after the establishment of the WTO in 1995, and especially with the mushrooming of regional trade agreements after the 1990s, the political economy of trade agreements began to look very different. The new-style agreements increasingly focused on domestic rules and regulations, such as intellectual property rights, investor rights, health and sanitary regulations, subsidies, and so on.

Unlike in the case of tariffs and quotas, there is no natural benchmark that readily allows us to judge whether a regulatory standard is excessive or protectionist. Different national assessments of risk—safety, environmental, health—and varying conceptions of how business should relate to its stakeholders—employees, suppliers, consumers, local communities—produce different standards, none obviously superior to others. In other words, regulatory standards are public goods over which different nations have different preferences. An optimal global governance scheme would trade off the benefits of expanding market integration (by reducing regulatory diversity) against the costs of excessive harmonization. But it is difficult to know where that optimal point may lie. Asking trade negotiators to perform this task adequately across a wide variety of policy domains seems unrealistic.

And this is before we allow for the political influence of internationally oriented special interest lobbies, which have played a critical role in these new domains, by shaping the formulation of global intellectual property regulations, investor arbitration clauses, banking standards and many others. Public information in the United States on lobbying for trade issues shows that pharmaceutical manufacturing firms and PhRMA (the industry association) top the list by a wide margin. Other significant contributors are auto manufacturers, milk and dairy producers, textiles and fabrics firms, information technology firms, and the entertainment industry. Labor unions such as United Steelworkers and the AFL-CIO, which are traditionally associated with protectionist motives, tend to lag considerably behind these industry-based groups.

These considerations suggest a different political economy model than the one economists have long been partial to. The domestic game that is played is not one between a free-trading government and protectionist interests, with international commitments serving to tie the government’s hand against protectionism. Rather, it is one where large international firms capture the international policy-making process to design global governance regimes in IPRs, banking, investment rules, etc., that are highly partial to their own interests. Unlike in the conventional model, the rent-seekers here are not the traditional protectionists. Instead, they are pharmaceutical companies seeking tighter patent rules, financial institutions that want to limit the ability of countries to manage capital flows, or multinational companies that seek special tribunals to enforce claims against host governments. In this setting, trade agreements serve to empower special interests, rather than rein them in.

Democracy-Enhancing Global Economic Governance

Whether international agreements can systematically alter domestic political equilibria in a desirable direction is a question with no clear-cut answer in theory. The recent evidence from trade agreements, reviewed briefly in the previous section, is not encouraging. Moreover, using external restraints to shape domestic policy has a certain cost in terms of democratic legitimacy: it reinforces nativist populists’ message of sovereignty being ceded to cosmopolitan technocrats. It should not be up to the “global community” to tell individual nations how they ought to weight competing domestic goals and priorities.

This does not preclude a global conversation over the nature of diverse benefits and harms to the parties. Such conversations can be helpful in reducing international misunderstandings about policy objectives, and sometimes in establishing new behavioral norms. When adverse economic spillovers are large, other countries may be able to convince governments that are the source of these spillovers to engage in better policies. International dialogue can enable some Coasian bargains to be struck when the losses incurred by other nations exceed domestic benefits.

The question I have tried to answer in this paper is: what are the circumstances under which countries should enter into binding international agreements? A somewhat different question relates to the form that these agreements should take—not just which policies should be covered, but also what types of domestic policy processes should be encouraged or discouraged. The article I mentioned above by Keohane, Macedo and Moravcsik 21 argues that multilateral agreements help democracies function better 22 . While I disagree with this conclusion as a general rule 23 , it is possible to turn their argument on its head and use it as a normative proposition (about how things should be) instead of a positive one (about how things are). Accordingly, we can envisage an alternative conception of global economic governance that directly targets potential domestic governance failures without presuming either that the appropriate national policies are known ex ante, or that global governance can have a significant impact. I call this democracy-enhancing global governance (DEGG), after Keohane, Macedo and Moravcsik 24 .

We can usefully distinguish DEGG from “globalization-enhancing global governance” (GEGG), which comes closer to the spirit of prevailing practice in the world economy today. Under GEGG, we can justify any and all external rules that restrict domestic policy autonomy if the result is to minimize transactions costs associated with national borders. Under DEGG, we would impose only those (mostly procedural) obligations that enhance domestic deliberation or are consistent with democratic delegation.

I have in mind procedural requirements designed to enhance the quality of domestic policy making. Examples of such requirements would be global disciplines pertaining to transparency, the broad representation of stakeholders, accountability, and the use of scientific/economic evidence in domestic proceedings. These procedural requirements would not prejudge what the end result might be—whether a country might impose a tariff, subsidy or any other “beggar thyself” policy.

Disciplines of this type are already in use in the WTO to some extent. The Agreements on Safeguards and Anti-Dumping specify domestic procedures that need to be followed when a government contemplates restricting imports from trade partners. Similarly, the SPS Agreement explicitly requires the use of scientific evidence when health concerns are at issue. Procedural rules of this kind can be used much more extensively and to greater effect to enhance the quality of domestic decision-making. For example, anti-dumping rules can be improved by requiring that consumer and producer interests that would be adversely affected by the imposition of import duties take part in domestic proceedings. Subsidy rules can be improved by requiring economic cost-benefit analyses.

We should not exaggerate the positive contribution such requirements can make to domestic decision-making. Consider, for example, Trump’s national security argument for hiking tariffs on steel and other imports. This was a classic beggar-thyself policy. WTO principles in this area are vague and remain largely untested in practice. On the one hand, the relevant text seems to open the door very wide by saying “Nothing in this Agreement shall be construed to prevent any contracting party from taking any action which it considers necessary for the protection of its essential security interests.” (Article XXI of the Marrakesh Agreement). On the other hand, in a recent ruling in a case not involving the United States, the WTO adopted the position that it can review national decisions in this area and judge their appropriateness. Predictably, the United States has criticized this decision 25 .

One can imagine more explicit rules about the process the United States (or any other country) must go through before the national-security case is established. For example, has the government prepared a public report, with input from economists and national security experts, which lays out the case in favor? Have the domestic opponents of the policy been given the chance to make the case against? Nevertheless, it is doubtful that any WTO approach would have made a difference to Trump’s trade follies. But at least it might have denied Trump (and other nativist politicians) the grounds for the habitual complaint that the WTO and other international bodies are trampling on national sovereignty.

At best, the light governance rules of DEGG I propose here can help somewhat. At worst, they do no harm. These rules entail soft disciplines for countries that already uphold democratic norms at home and that would benefit from the additional reinforcement that comes from international fora. But it is also possible to envisage harder disciplines, where countries get the full benefits of trade agreements only to the extent that they live up to democratic commitments. Such requirements already exist in some trade agreements (e.g., the U.S. system of trade preferences for low-income countries).

Concluding Remarks

International agreements are contracts into which nations freely enter. And since they are voluntary contracts, there would seem little reason to question them on the basis of loss of national autonomy or democratic legitimacy. But this approach begs the question of why states enter into such contracts in the first place. To have democratic legitimacy, international agreements have to pass political and economic tests: they must produce broad benefits and be consistent with democratic delegation (impose restraints that enhance democratic functioning). By developing principled arguments for global economic governance, we can clarify the set of circumstances under which such contracts are broadly desirable—as well as distinguish these circumstances from instances where the contractual nature of international agreement is used as a cloak to hide the privileging of particular special interests.

When, for example, U.S. trade negotiators obtain TRIPS concessions from another country in return for opening the U.S. market to that country’s exports of garments, they effectively trade-off gains to Big Pharma against concentrated losses on some segments of the domestic labor market. How do we think about the appropriateness of such a contract? Fundamental economic arguments of the type I have examined here are critical for supplying appropriate justifications. In the presence of BTN/GPG considerations, the contract could be win-win. In their absence, what superficially appears to be win-win—a mutual exchange of market access—is essentially a policy that induces a first-order redistribution at home. International agreements would have more democratic legitimacy at home in the first instance than in the second.

Of course, even in BTN/GPG cases, there is no guarantee that democracies will solve their domestic political problems and reach appropriate international bargains. The U.S. withdrawal from the Paris climate agreement is a notable example. Such problems are yet another reflection of one of the key arguments in this paper: most policy mishaps in the world economy today occur due to failures of national governance, not due to lack of international cooperation 26 . To take yet another American example, Trump’s national security tariffs were bad policy not because they harm certain other nations; they are bad policy because they impose substantial costs directly on the U.S. economy.

Conventional wisdom on global governance relies on international coordination failures arising from global public goods or beggar-thy-neighbor policies. When the troubles originate with beggar-thyself policies instead, or legitimate grounds for diversity in economic policies, this perspective is no longer helpful. For such circumstances, we need to update our thinking. We need to adopt a different approach to global cooperation, one that respects the policy space of nations and targets democratic decision-making norms instead of one that emphasizes the harmonization of policies or the removal of (real or perceived) trade barriers.


  1. Dani Rodrik is with the Harvard Kennedy School. The author is grateful to the World Bank for financial support and to Robert Cook, Robert Keohane, Robert Staiger, and especially Harlan Grant Cohen for very useful feedback.
  2. Editor’s note: this is a revised version of a paper prepared at the World Bank ABCDE conference on June 17–18, 2019 in Washington, DC and subsequently published as Rodrik, D. 2020. “Putting Global Governance in Its Place”, The World Bank Research Observer, vol. 35, no. 1. As such, the ideas contained in this paper were developed at a time when Donald Trump was still the U.S. President, and Covid-19 was not yet a global pandemic.
  3. See, Friedman, T. L. 2019. “Has Our Luck Run Out?” New York Times, April 30, 2019.
  4. See, Shafik, N. 2013. “Smart Governance: Solutions for Today’s Global Economy.” International Monetary Fund, December 3, 2019.
  5. See, Eichengreen, B., and R. Baldwin. 2008. “What the G20 Should Do on November 15th to Fix the Financial System.” Vox, November 10, 2008.
  6. See, 1.    See U.S.-China Working Group on Trade Policy (2019) for a statement along these lines, and an approach that follows the ideas in the paper.
  7. There is a rich body of analytical literature on the economics of trade agreements, which is complementary to my discussion here. See, Bagwell, K., and R. W. Staiger. 2001. “Domestic Policies, National Sovereignty and International Economic Institutions.” Quarterly Journal of Economics 116 (2): 519–62; Bagwell, K., and R. W. Staiger. 2004. The Economics of the World Trading System. Cambridge, MA: MIT Press; Frieden, J., M. Pettis, D. Rodrik, and E. Zedillo. 2012. After the Fall: The Future of Global Cooperation. Geneva, Switzerland: Geneva Reports on the Global Economy International Center for Monetary and Banking Studies.; Grossman, G. 2016. The Purpose of Trade Agreements. Princeton: Princeton University; and the references therein. See also Gallagher, K. P., and R. Kozul-Wright. 2019. ANew Multilateralism for Shared Prosperity: Geneva Principles for a Global Green New Deal. Boston, MA: Boston University Global Development Policy Center and UNCTAD, for an articulation of a new set of principles for the future of multilateralism, and Cohen, H. G. 2019. “What is International Trade Law For?” American Journal of International Law 113 (2): 326–46, for an argument for “re-embedding” international trade law in domestic policy priorities.
  8. For a formalized version of the arguments developed in this paper, see Rodrik, D. 2020. “Putting Global Governance in Its Place”, The World Bank Research Observer, vol. 35, no. 1.
  9. See, Johnson, H. G. 1953. “Optimum Tariffs and Retaliation.” Review of Economic Studies 21 (2): 142–53.
  10. See, U.S.-China Trade Policy Working Group.  2019.  U.S.-China  Trade  Relations:  The  Way  Forward. October 27, 2019. joint_statement_2019_0.pdf .
  11. A commitment problem arises because each country would still like to deviate from the cooperative equilibrium and resort to BTN policies. The issue of commitment also raises the question of why a cooperative outcome could not be obtained through repeated-interaction incentives in dynamic games, instead of relying on an international agreement or organization such as the WTO. Formal governance structures may have an advantage in that they allow for coordination when there are multiple equilibria to select from and provide information on compliance in settings with many players (see discussion in Grossman, G. 2016. The Purpose of Trade Agreements. Princeton: Princeton University).
  12. See, Tørsløv, T., L. Wier, and G. Zucman. 2018. “The Missing Profits of Nations.” NBER Working Paper No.24701 (August).
  13. There is also a wide range of circumstances with imperfect competition where governments may want to shift rents from foreign to domestic firms. Policies used in such cases sometimes look like BTN policies, but the presence of imperfect competition means that their global efficiency consequences can be quite different from standard BTN policies. For example, when a country subsidizes its domestic oligopolist (say Airbus) to shift rents from a foreign oligopolist (Boeing), the rest of the world benefits through lower prices.
  14. A possible exception arises in the case of truly predatory pricing: subsidies may enable the home firm to drive foreign competitors out of business and subsequently exert monopoly power on world markets. But such cases are rare (and would fall under BTN policies discussed previously). Current trade rules do not single out predation (as they should).
  15. I am grateful to Robert Cook for reminding me of this argument.
  16. I provide a concrete example in my proposal for an anti-social dumping clause in Rodrik, D. 2019. 2019. “Policy Brief No. 9: Towards A More Inclusive Globalization: An Anti-Social Dumping Scheme.” Economics for Inclusive Prosperity. This proposal seeks to marry national autonomy in upholding domestic social bargains with the global procedural restraints discussed below.
  17. In some areas, there may well be well-established global norms. (I will treat democracy as one such global norm later in the paper when I discuss democracy-enhancing globalization.) But in other areas, including especially the treatment of labor markets, practices obviously differ.
  18. See, Keohane, R. O., S. Macedo, and A. Moravcsik. 2009. “Democracy-Enhancing Multilateralism.” International Organization 63 (Winter): 1–31.
  19. See, Rodrik, D. 2008. “Second-Best Institutions.” American Economic Review 98 (2): 100–4.
  20. Rodrik, D. 2018. “What Do Trade Agreements Really Do?” Journal of Economic Perspectives 32 (2): 73–90.
  21. See, Keohane, R. O., S. Macedo, and A. Moravcsik. 2009. “Democracy-Enhancing Multilateralism.” International Organization 63 (Winter): 1–31.
  22. At least one of the authors (Keohane) seems to have changed his mind subsequently (see Colgan, J. D., and R. O. Keohane. 2017. “The Liberal Order is Rigged: Fix it Now or Watch it Wither.”Foreign Affairs (May/June): 36–44).
  23. See, e.g., Rodrik, D. 2018. “What Do Trade Agreements Really Do?” Journal of Economic Perspectives 32 (2): 73–90.
  24. See, Keohane, R. O., S. Macedo, and A. Moravcsik. 2009. “Democracy-Enhancing Multilateralism.” International Organization 63 (Winter): 1–31.
  25. This is a case involving Russia’s transit restrictions affecting Ukraine. See
  26. The word “today” is important in this sentence. It is possible to envisage situations where failure in international cooperation plays a much more significant role. Trade and macroeconomic policies during the 1930s provide an example. Similarly, a future world where statist/nationalist governments pursue a wide range of BTN policies is not unimaginable.
voir le planfermer
citer l'article +--

citer l'article


Dani Rodrik, The Right Scope of Global Governance and Democracy Enhancement, Aug 2021, 36-44.

notes et sources +