Reflecting on populism and the economics of globalization
In this commentary, we take up two key elements of Dani Rodrik’s analysis of the economic underpinnings of populist politics. We focus on the links between globalization (especially trade integration) and populism and the role of global institutions (notably trade agreements) in both generating the economic pressures to which populist politicians are responding and in constraining the ability of governments to deal with adverse distributional effects of these pressures. We argue that it is important to distinguish between trade shocks and trade agreements; that the role of both is given too much weight relative to the effects of financialization and international capital flows, migration, and technological change; and that deepening international cooperation (global governance) can – and should – be part of the supply response to populism.
Keywordspopulism globalization deep integration
Dani Rodrik has an outstanding multi-decade track record of high-impact scholarship on globalization, political economy, and development. One strand of his research has centered on the distributional dimension of globalization and its consequences for the political sustainability of international integration of capital and product markets. A recurring theme of his work has been the importance of national “policy space”: that governments be free to adopt policies that they deem to be appropriate to achieve national economic development goals – and to experiment – without being constrained by international institutions. He is skeptical regarding the welfare benefits of international initiatives aiming at so-called deep integration: agreements that go beyond reducing barriers to trade at the border to include disciplines on domestic policies. These themes recur in his discussion of populism and the economics of globalization that is included in this issue of the Journal of International Business Policy (Rodrik, 2018).
There is much to like about his analysis of the economic dimensions of the rise of populism. His insights into why populism is more likely to take the form of left-wing or right-wing politics and the role that liberalization of capital flows has played as a driver of inequality and economic hardship (crises) are particularly salient. In this comment, we take up two elements of his argument: (i) the link between trade integration and populism, and (ii) the role of trade agreements in generating these pressures and constraining the ability of governments to manage adverse distributional dynamics. We will argue that the explanatory power of trade as a driver of populism is limited and that it is important to distinguish between trade shocks and trade agreements. The trade shocks that have occurred in recent decades are primarily the result of autonomous actions (policy reforms) by governments and technological developments that are driving structural change. Trade agreements have not been a major driver in our view. Finally, we are more positive than Rodrik regarding the role international cooperation (global governance) can play in helping national governments manage the potential downsides of globalization and address adverse income distributional effects. Sustaining and deepening international cooperation and improving global governance are key dimensions of the political supply response needed to address the rise of populism.
Globalization and Populism: What is the Connection?
Rodrik begins by seeking to tie current populism to standard trade theory, with the implication that the emergence of anti-globalist populism was predictable. His analysis is presented in terms of a comparative static exercise based on the standard general equilibrium model of international trade. This assumes perfect competition in perfectly functioning markets, and the analysis proceeds by comparing two equilibria, before and after a change in trade prices, after all relevant economic adjustments have occurred (i.e., the “long run”). This is an ideal framework for thinking about the effects of different policy regimes because it abstracts from adjustment issues.1 It reveals that, in the long run, although the gains will outweigh the losses, some people will experience real losses from trade liberalization. The overall gains from international integration come from the reallocation of resources from low to higher productivity activities, and restructuring of activities that only were competitive because of policies making it more costly for domestic consumers to buy foreign-produced goods and services.2 In the case of high-income countries, losers are likely to be workers and households that derive income from (are employed in) relatively highly paid, relatively unskilled activities in which foreign countries have comparative advantage.
This framework is not very good for analyzing the short-run adjustments to a change in the economic environment resulting from trade liberalization, domestic policy reform, or a change in the productivity of a trading partner. The 2-good \(\times\) 2-factor version of the Stolper–Samuelson theorem serves the very useful purpose of illustrating that a change in trading conditions (or anything else that might change relative prices, e.g., technological change, changed regulatory conditions, changes in tax policy, etc.) can have an unambiguous effect on returns to a particular factor of production (raising the real return to one factor and lowering the return to the other). Away from the case where the number of goods and factors are both equal and small (i.e., 2) things become more complicated. As Rodrik says, we can prove that some factor loses, but he fails to note that in the high-dimensional case, we cannot generally identify the loser. The result is local in a mathematical sense, in that the loser under one set of parameters will not generally be the loser under another. Moreover, the identity of the loser depends sensitively on the actual change in the vector of commodity prices. Not surprisingly, it has proven very difficult to pin down empirical Stolper–Samuelson effects (see e.g., Slaughter, 2000). Perhaps more importantly in this context, the link between Stolper–Samuelson like results and political beliefs and actions is, at best, very weak (Hiscox, 2006; Mansfield & Mutz, 2009).
The point here is not that trade cannot have an effect on politics, just that the long-run trade models developed for comparing policy regimes, and whose empirics suggest modest long-run effects, are not terribly useful for understanding the emergence of anti-globalist populism. By contrast, the transitional costs of adjusting to changes in the trade policy environment can be quite significant, while geographically and sectorally quite local (e.g., Autor, Dorn, & Hanson, 2016; Hakobyan & McLaren, 2016). There are both ethical and political reasons for responding to these losses. From an ethical perspective, the gains from trade associated with liberal trade policy are actually produced by moving inputs from less to more efficient sectors. However, another way of characterizing “moving inputs” is “transitional unemployment.” We have already noted that these costs can be locally quite significant, but it is precisely the “trade displaced workers,” i.e., the people who are paying the short-run costs of adjustment, who are producing the gains from trade (this is also emphasized by Rodrik). Virtually, all coherent ethical systems would see this as a warrant for some form of assistance. As a practical political matter, if these costs are significant, not only does this create a class of people with a common interest in resisting the trade shock, but the manifest inequity creates a basis for wider mobilization. As a result, virtually all textbooks point to the need for compensation mechanisms for market opening (liberalization) to be Pareto superior to the status quo ante.3 Estimates of the extent to which formal trade adjustment assistance, and other transfer programs, offsets these losses suggests that the offset is substantially less than complete (Autor et al., 2016, Section 4.4). It is hard to link the rise of anti-global populism to trade adjustment costs, but there certainly may be a link to more generalized economic hard times. If this is the case, general welfare state policies will play a much more significant role than trade adjustment assistance.
Rodrik contends that major drivers of the rise of populism are the uncompensated adjustment costs and redistributive effects of economic globalization.4 A problem with this argument is that the evidence of these costs is associated with two specific events: NAFTA and the “China shock” (accession to the WTO). For all of the rhetoric surrounding NAFTA, it is hard to see how locally significant transition effects (Hakobyan & McLaren, 2016) that resulted in overall effects that were modestly positive (and tended to be underestimated at the time – see e.g., Trefler, 2004; Kehoe, 2005; Caliendo & Parro, 2015), could have driven populism more than a decade and a half after the adjustments had occurred. The China shock, by contrast, surely was large and, again locally, substantial (Autor et al., 2016), but it was also global, as has been the rise of anti-global populism. However, our reading of recent events is that most of the anti-global rhetoric was not targeted on China.5
One reason for this is that there have also been significant increases in employment in conjunction with reductions in employment in manufacturing. These increases have occurred in services sectors and in export-oriented activities. In the case of the US, in the aggregate there has been large-scale job creation that more than outweighs the losses in manufacturing employment. Feenstra and Sasahara (2017) calculate that trade with China increased labor demand by a total of 1.7 million jobs, much of it in services. As a services economy, overall employment growth in the US has largely been concentrated in services sectors, a factor that has tended to be ignored or neglected in the analysis of labor economists of the effects of greater trade on manufacturing employment.
More importantly from the perspective of understanding the material basis of contemporary populism is that the China shock is historically unique – a function of a fundamental change of policy in China.6 But if this is the source of anti-globalism, it is now largely over. The world economy has experienced the shock and most of the adjustment is nearing completion. Conditioning future trade policy on the China shock is essentially a category mistake. Furthermore, raising trade barriers will have the same sort of transitional effects, while producing a reduction in national income (i.e., as a result of moving to a more inefficient long-run equilibrium). Presumably, that would also produce a political response that might take the form of anti-globalism, but the material foundation for those politics would be even shakier than those for current populist politics.
Abstracting from the distinctive China shock, “globalization,” however defined, is a complex phenomenon involving trade in goods and services, reorganization of international production structures, cross-border flows of finance and information (knowledge), and movement of people of a wide variety of sorts (refugees, skilled workers, unskilled workers, family members, etc.). As a result, the diverging experiences and situations prevailing across countries that have seen a rise in populism indicate a need to “unpack” different dimensions of globalization. Trade agreements loom large in Rodrik’s work, and have been a significant element in anti-globalization on the US left (Bernie Sanders). However, in both the US and Europe, most voters are far more concerned with migration and issues such as financialization. In this regard, it is important to note that the politics of migration are almost completely unconnected to economic consequences (which are generally estimated to be quite modest – e.g., Card, 2009). The role of identity politics is very clear in this case and, we would suggest, this is also true of anti-trade politics.
National policies towards migration, capital flows, and regulation of the financial sector are not covered in a serious (binding) way in most trade agreements. The EU is very much an outlier in this regard. Europe is certainly important for understanding drivers of populism and the consequences of adopting a common currency and allowing for freedom of movement of people. However, one needs to be careful not to extrapolate from the EU as it is the only extant true deep integration initiative. The WTO – the bad guy for many populists (e.g., Trump) – is not a deep integration agreement.7 It does not cover migration policies or impinge on monetary or exchange rate policy. It provides extensive latitude for governments to use trade policies to respond to adjustment pressures if they desire to. The same is true of most preferential trade agreements.
Compensation of short-run losses from international integration is important. In this we wholeheartedly agree with Rodrik. It is essential to note, however, that by our analysis this is a matter for domestic policy – welfare state income insurance and appropriately structured active labor market policy, not trade activism. There is no need for compensation to take the specific form of trade adjustment assistance (and certainly no warrant for a trade policy response). Since, as we suggest above, it is far from clear in most cases that trade is, in fact, the major source of adjustment pressure, general welfare state and active labor market policies are called for. The decline of these instruments, especially in the US, is only very indirectly related to globalization, as their greater robustness in Northern Europe suggests. Furthermore, we would take issue with the argument that compensation is very costly or is impeded by time inconsistency and credible commitment problems. The European experience illustrates that compensation – embedded in general social policies – is certainly economically and politically feasible; the empirical literature on the magnitude of aggregate adjustment costs/income losses suggests that the trade dimension of compensation is likely to be a relatively small part of the overall losses imposed on workers affected by job losses.
From the perspective of understanding the extent to which populism is due to international integration, the prevalence of compensation mechanisms (social welfare/safety nets) in Europe is important, given the rise in populism there. It implies that the existence or not of compensatory programs cannot be the only explanation for the rise in populism that has occurred since the mid-2000s. In Europe, the rise of populism is associated with a mix of migration (because the EU includes freedom of movement of workers) and narratives by national politicians that “Brussels” (the EU) constrains their ability to pursue policies that would increase growth, create employment, etc. The boom and bust following the introduction of the euro, the loss of national monetary policy and the exchange rate instrument associated with monetary union, and rejection by creditor countries of debt restructuring in favor of austerity policies played a clear role in the rise in populism in Europe post-2008. This is recognized by Rodrik in his analysis. But the fact that trade has not played much of a role is important for analyzing the potential drivers of populism and considering policy responses because few trade agreements constrain the ability of countries to restrict migration or to use monetary policies and maintain flexible exchange rate regimes. Again, the EU is a distinct case that offers important lessons for the design of deep integration initiatives, but trade is not a significant part of the story of the rise in populism in Europe.
The New Political Economy of Globalization
After spending the first part of the paper seeking to forge a connection between the material consequences of globalization and the rise of anti-global populism, the second part of Rodrik’s article turns to a more explicitly political account. The fundamental claim is that a widespread belief in the “unfairness” of current globalization underlies a more permanent opposition to those arrangements. Rodrik then argues that the appropriate response to this problem is to reconstitute the international economy with a stronger emphasis on sovereignty, defined as assuring (reclaiming) the ability to pursue distinctive national policies (especially progressive programs). There is some overlap here with the arguments of populist politicians (stripped of the racist anti-immigrant appeals)8 that the cause of our problems is foreign, so the solution is to take action to safeguard the national interest – whether this is defined in terms of jobs, social norms or national security – by rolling back liberalization and increasing discrimination against foreigners, whether the products they produce or the people themselves (closing the border to migrants, not accepting refugees, making it much more difficult and costly to obtain visas, etc.).
For Rodrik, the perception of foreign unfairness acts as a political accelerator of such views. There are two problems with this part of the analysis. First, foreign unfairness, relative to domestic policy, is hardly a fact. Consider Carrier. Consistent with Rodrik’s argument, in 2017 President Trump manufactured quite a bit of outrage about Carrier moving production to Mexico. What is interesting is that Carrier closed production facilities in upstate New York in 2004, moving production to right to work state Georgia. Right to work states permit restrictions on labor organization that were not legal in New York, but this was seen as a victory for market-friendly policy. Similarly, most US citizens would probably not find a decision by the EU to restrict imports from the US as a result of weak labor or environmental standards a reasonable policy. Thus, and this is the second point, the politics of fairness are opportunistic. It is not that fairness is a sham. Rather it is a powerful tool precisely because we voters respond strongly to shared perceptions of unfairness. The point is that such politics are poorly rooted in material fact. A politically effective claim of foreign unfairness may work in one setting and completely fail in another.
The presumption that globalization, or at least perceptions of the fairness of globalization, is a major driver of the current malaise raises the question of the relative significance of globalization, technological change, (lack of) national regulatory responses, weakening of unions, erosion of welfare state institutions, or any of the myriad other factors that might materially affect political preferences and perceptions of fairness.9 These are not independent factors, but (except as a matter of ideology) there is no clear causal structure linking them. Any claim asserting the dominance of any of these factors, and globalization in particular, is not well-founded. It does seem clear that, given the combination of factors at work, there is no change in trade policy that will increase employment and/or the average wages of unskilled workers without seriously damaging the overall economy.
A central political-economic claim in Rodrik’s analysis is that (hyper-)globalization threatens the sustainability of democratic capitalism. The intuition is Rodrik’s political trilemma of the world economy.10 Unlike Robert Mundell’s famous open economy trilemma, Rodrik’s trilemma is rhetorical, not analytical. Mundell (1957) shows, in the context of a well-specified open economy, Keynesian model, that fixed exchange rates, free capital mobility, and an independent macroeconomic policy targeted on a domestic policy goal (e.g., unemployment) are inconsistent. Note that these are all actual policy goals and that their inconsistency is derived from a plausible model of the underlying economy. By contrast, none of the elements of Rodrik’s trilemma (hyperglobalization, democracy, and sovereignty) are policy goals in quite the same way. Democracy and sovereignty are structural facts that are both seen, correctly, as essential elements of the political economies of most core states in the world economy, but surely “hyperglobalization” is not a goal. These three elements of Rodrik’s trilemma are definitionally linked: hyperglobalization constrains sovereignty; and sovereignty is essential to democracy; so, if sovereignty and democracy are to be saved, which Rodrik clearly sees as essential (as do we), his analysis says that hyperglobalization must go. This leads to the conclusion that “…the big challenge facing policymakers is to rebalance globalization so as to maintain a reasonably open world economy while curbing its excesses. We need a rebalancing in three areas in particular: from capital and business to labor and the rest of society, from global governance to national governance, and from areas where overall economic gains are small to where they are large.”
Questions can be raised regarding all three of these prescriptions. Regarding the first point, an underlying presumption is that capital (businesses) are disconnected from labor and the rest of society. This is not the case. Outside the Heckscher–Ohlin–Samuelson model and the long-run environment to which it applies, politically relevant winners and losers from globalization do not generally break down along class lines. Rather, as the research cited by Rodrik suggests, occupations, sectors, and regions are more relevant units. These are sufficiently clear that tax-and-transfer policies more directly connected to the actual economics are quite feasible. Turning to the third point, it may well be that small gains in some areas are relatively easy to obtain, i.e., at low political cost and without negative repercussions for incomes/jobs, whereas areas with large gains may confront major political costs/constraints. Some effort to identify these areas may well pay dividends.
Our main concern is with the second area where Rodrik calls for rebalancing: a shift away from global governance towards greater (and hopefully better) national governance. The challenge of putting in place better domestic policy is not conditional on reducing or attenuating the focus on global governance. To the contrary: there is an opportunity to reduce the burden placed on domestic policy through international cooperation. In part, this may be achieved by reducing the negative externalities imposed by national policies. But global governance mechanisms may also enhance the quality and effectiveness of domestic policies by identifying good regulatory practices and encouraging learning. Rodrik takes the view that there is already too much global governance and that it is not needed (i.e., is undesirable). We agree that better domestic policies are critical but would argue that global governance is desirable and while difficult, is feasible and should be pursued.
The call for rebalancing away from global governance emerges organically from Rodrik’s trilemma, but given our criticism of both the emphasis on globalization and the rhetorical nature of the trilemma, we find this prescription misplaced. In recent work (Hoekman & Nelson, 2018), we offer an alternative analysis rooted in the well-understood tension between capitalism and democracy.11 Because this tension is inherent, it is constantly being renegotiated politically. This is the key insight of Polanyi’s “double movement” (Polanyi, 1944, Chapter 11): capitalism and democracy are in a dynamic tension that must be managed. We sketch a structural relationship between capitalism, democracy, and the nation state, in which the state plays a central role in managing the tensions between civil society (the social foundation of democracy) and capitalism. Thus, the liberalizing trend associated with deregulation, de-unionization, liberalization of global economic relations, etc. has been increasingly met by the mobilization of what Polanyi called forces of “social protection.” These politics are conjunctural – they vary across time and space as a function of the specifics of the moment.
The driving economic force of the current conjuncture is the transition to a post-industrial economy in the core nations of the world economy. This involves fundamental changes in the structure of production involving the rapid growth of small, efficient service firms and the expansion of global production strategies (“global value chains”) of large multinational enterprises. This has often meant substantial change in what jobs are done where, and how those jobs are organized. Not surprisingly, it has produced widespread concern that is easily mobilized politically. Through much of the post-war era, political mobilization was organized in terms of a left–right dimension defined primarily in economic terms, but beginning well before the current populist boom, a new, broadly cultural, dimension has emerged as a major structuring principle (Hooghe, Marks, & Wilson, 2002; Kriesi et al., 2012). Faced with migration pressures, financial crises, and the China shock, the extreme right has been able to mobilize using these issues as an essential part of this cultural dimension.
Clearly, large changes in the economy associated with post-industrialization challenge the political supports of democratic capitalism. Globalization is a fundamental part of this story. A key feature of current globalization is what Baldwin (2016) has termed the “second unbundling” – the emergence of global production structures anchored on the US, Europe, and Japan, with China in a special relationship to all three of these regional hubs. Where the earlier version of globalization could be relatively easily managed through a combination of national economic policy and relatively straightforward multilateral controls on trade distorting policies, complex production relations distributed across many countries rely on the kind of predictable legal and political structures that played an essential role in the development of national capitalism. Rodrik’s argument against global governance in favor of “subsidiarity” assumes few international spillovers, whereas interdependence associated with generalized global production in fact calls for greater global governance. This is a challenge, but not an impossibility.
Rodrik is skeptical as regards the benefits of international cooperation and agreements. Elsewhere he argues that global governance is a “false promise” (Rodrik, 2017: 206) because civil society is national and so are nation states. In the current article, he argues that “the world economy does not really need global governance to be managed appropriately [because] [m]ost failures in the world economy are rooted in failures of domestic governance.” While he recognizes that (better) global governance could help, on balance the downside risks are stressed. The bottom line is that “…in rebalancing globalization toward national governance, we must understand that the best contribution global arrangements can make is to make the nation state work better, not to weaken or constrain it. Correspondingly, the appropriate role for global institutions is to enhance key democratic norms of representation, participation, deliberation, rule of law, and transparency – without prejudging policy outcomes or requiring harmonization.”
We agree with the thrust of this but disagree that constraining the nation state may not be desirable. While democratically legitimate trade policy must be fundamentally rooted in national concerns, the current form of trade globalization centered on extensive specialization of firms connected through regional and global value chains literally requires international cooperation. This does not imply binding one-size-fits-all rules that constitute a threat to national sovereignty and democratic legitimation. Rather, what is needed is international regulatory cooperation that reduces risks of catastrophic failures and enhances the ability of regulators to attain the health, safety, and social standards that have been established by their respective national polities. The same applies to the regulation of international finance, cooperation to address erosion of the tax base (taxing mobile firms), or helping governments deal with the negative effects of competition in the granting of investment incentives. Constructing such arrangements is not easy, but is essential. There are in fact many successful examples of international regulatory cooperation that is deemed to be desirable by participating governments (see e.g., Hoekman & Sabel, 2017). Thus, the challenge of putting in place better domestic policy is not conditional on reducing or attenuating the focus on improving global governance. To the contrary: there is an opportunity to reduce the burden placed on domestic policy through international cooperation. In part this may be achieved by reducing the negative externalities imposed by national policies – which implies agreement among sovereign states to refrain from using certain policies and/or to adopt what has been agreed by the countries involved to constitute good practices to achieve a given policy objective.
Rodrik recognizes that trade agreements are not necessarily the major driver of globalization and the adjustment costs and losses that have been incurred by some groups in society. He points in particular to the effects of financial globalization and the attendant volatility and periodic major crises. He also notes the importance of immigration as a factor driving support for populism in Europe. The multidimensional nature of the factors that have supported a rise in populism suggests a multidimensional response is in order. However, the main focus of Rodrik’s discussion of policy is on product market integration, arguing that deeper versions of trade agreements that span regulatory policies undermine sovereignty and are instruments through which powerful business groups and associated elites generate rents at the expense of consumers and workers and undercut the policy space of (future) governments (e.g., Rodrik, 2017). These arguments exaggerate what is in fact done in most trade agreements, attribute specific measures agreed by governments to trade agreements when these do not (and need not) encompass the instruments that are objected to, and neglect the possibility of learning and revising the substance of agreed rules.12
For example, investor-state dispute settlement (ISDS) provisions have for a long time been part and parcel of bilateral investment treaties (BITs). More recently they have begun to be incorporated into trade agreements that cover investment-related policies, but as illustrated by recent developments in the EU, there is no need for trade agreements to include BIT-type ISDS. The fact that these matters came to be included in recent trade agreements raised their public profile, and the resulting contestation and debate are leading to changes in the approaches taken in this area. The Comprehensive and Progressive Agreement on Trans-Pacific Partnership that was negotiated and agreed by eleven of the original signatories of the Trans-Pacific Partnership (TPP) following the decision by President Trump to withdraw from that agreement does not extend to implementation of many provisions on protection of pharmaceutical products that were included at the behest of the US.
These developments illustrate two things: (i) trade agreements are negotiated and thus reflect a balance of interests that may well result in inclusion of provisions that benefit certain industries/interest groups that are welfare-reducing on a stand-alone basis; and (ii) learning and adaptation in the structure and substance of agreements occurs. We agree with Rodrik that the first dimension is objectionable, although presumably the governments involved in the TPP were of the view that whatever the US was offering more than offset the costs of IPR provisions pursued by the US. These are trade-offs that each national government and parliament must consider. Political economy dynamics in every country will determine whether negotiated agreements are approved and presumably this is a matter that should be left to each government/parliament to determine. Rather than reject the possibility of political renegotiation and learning, we would emphasize the importance of putting in place mechanisms that support such dynamics and that empower parliaments and national polities to better assess the costs and benefits of specific instances of international cooperation.
Arguing that the best response to populist pressures and tensions is a democratic rebalancing to restore national autonomy seems to us to do too little to recognize that deeper international cooperation can – and in our view should – be part of the supply response to the drivers of populism. Better global governance can come in many forms. One element we would stress is more multilateral engagement by China and other middle-income countries that have benefitted from the multilateral trading system and an open world economy. If they do not, the global trade regime that has helped underpin the massive increase in global incomes and the attendant reduction in global inequality and absolute poverty in many countries may unravel. The associated increase in the exploitation of national policy space can easily be to the detriment of most, if not all, countries. While China is now pursuing a more activist external policy that aims at least in part to support its trading partners, e.g., the Belt and Road Initiative, more engagement in the WTO and other multilateral mechanisms is needed to strengthen global governance of trade and production. The desirable design features of such governance are beyond the scope of this comment, but certainly extends beyond traditional trade agreements to span international financial flows, regulation of the financial sector, and taxation of international business. There is no one-size-fits-all model, and experimentation should be encouraged – e.g., among subsets of countries on a plurilateral basis (Hoekman & Nelson, 2018) – but improving global governance is part of the solution.
This class of model also abstracts from such things as entry by more efficient, and exit of less efficient, firms, technological change associated with more competition and access to better intermediate goods and increased varieties of goods. All of these increase gains from trade. Empirical research suggests that all these elements played a role in underestimating the gains from integration agreements such as NAFTA (Trefler, 2004; Kehoe, 2005; Caliendo and Parro, 2015). Models incorporating such factors reduce the relative magnitude of negative distributional effects of reducing barriers to trade.
Much of the rhetoric of anti-globalization emphasizes losses to specific people, so it is important to remember that the logic that identifies those losses also identifies gains that outweigh those losses. While it is natural to worry about losers from any particular change, it is a standard issue in welfare economics that, in situations with winners and losers, there is no way around making explicit welfare judgments. In this case, a preference to reverse policy to protect losers amounts to a willingness to damage winners. We might think that this is supported by some form of prioritarian reasoning that privileges the worst off in society (à la Rawls or Dworkin), but it is not at all clear that the main losers are the worst off. In fact, the worst off in society are surely the beneficiaries of lower prices, while the losers were making premium wages in declining industries.
We tend to discuss trade adjustment assistance in terms of workers, but losers will also include capital owners of affected firms that confront greater import competition post-liberalization (and more generally experience greater competition in third-country markets as a result of policy reforms by foreign countries, e.g., discriminatory trade agreements). Compensation arguments generally refer to individuals (income losses incurred by workers from job loss), not the firms negatively affected by trade liberalization and the associated loss of rents, though there are trade adjustment assistance programs targeting firms (Hornbeck, 2013).
Rodrik supports his analysis at this point with his political cost–benefit ratio (PCBR). While a useful classroom exercise, this has never been shown to map to actual politics of trade, and certainly not in the case of the recent rise of populism. More important, the PCBR is very much a partial equilibrium exercise.
China’s re-integration into the world economy led to a massive, unprecedented increase in the global labor supply. This was one-time event that has not yet played out fully, but it is worth noting that notwithstanding the size of the shock to effective global labor supply, China is simply a blip in the longer term trend in the decline in the share of manufacturing in total US employment (Baily & Bosworth, 2014).
The driver for the growth of China was an extensive program of autonomous, unilateral, reforms in a very broad range of policy areas, many of which had nothing to do with trade or the WTO. The resulting adjustment costs for manufacturing workers in the US and Europe arguably would have happened with or without China’s accession to the WTO.
For us, deep agreements imply binding commitments to pursue specific forms of policy – what Tinbergen (1954) called positive integration. This contrasts with shallow agreements that discipline the use of border policies. All trade agreements include numerous holes and loopholes permitting governments to temporarily cease to apply such agreed disciplines when needed for domestic policy purposes.
This is somewhat tricky since it is precisely the racist anti-immigrant appeals that are the political backbone of the current right-wing populism that is ascendant in the wealthy West. See, e.g., Müller (2016).
One dimension of fairness that is discussed by Rodrik and that is quite salient in our view is the argument that integration initiatives increasingly include a focus on domestic policies that reflect social values and preferences. This raises concerns by voters that cooperation may erode national standards. Efforts to reduce the transaction costs associated with differences in domestic policies that are regulatory in nature make international cooperation to reduce such costs more contested than standard trade liberalization. This is true especially when the locus of attention concerns nonpecuniary externalities – associated with issues such as foreign labor standards and protection of the environment. This “policy interface” dimension of globalization may help explain why populism is rising in countries with strong welfare states and that have seen aggregate real incomes (GDP) grow substantially over time – e.g., Europe. As we argue below, this calls for more and better international cooperation – global governance – not less.
The trilemma does not appear in Rodrik’s article but it is important in understanding his thinking on global governance. A clear and complete presentation can be found in Rodrik (2011, Chapter 9).
This tension has been analyzed for virtually as long as capitalism and democracy have been attributes of the modern political economy. The tension is often commented on in moments of crisis. Thus, for example, the great depression called forth such comment from both the right (Schumpeter, 1942) and the left (Polanyi, 1944).
Abstracting from the important case of the EU, policy space is nowhere near as constrained by international trade agreements as tends to be claimed by many critics – or proponents. A basic question is whether discrimination against foreign products is an optimal policy to achieve a domestic objective, and when it may be an appropriate second-best instrument because first-best policies are not available.
We are grateful to Sarianna Lundan, Laura Seelkopf, and Ari Van Aasche for their helpful comments on an earlier draft.
- Baldwin, R. 2016. The great convergence: Information technology and the new globalization. Cambridge, MA: Harvard University Press.Google Scholar
- Feenstra, R., & Sasahara, A. 2017. The “China Shock,” exports and U.S. employment: A global input–output analysis. NBER Working Paper 24022.Google Scholar
- Hoekman, B., & Nelson, D. 2018. 21st century trade agreements and the Owl of Minerva. CEPR Discussion Paper 12620.Google Scholar
- Hoekman, B., & Sabel, C. 2017. Trade agreements, regulatory sovereignty and democratic legitimacy. EUI Working Paper RSCAS 2017/36.Google Scholar
- Hornbeck, J. F. 2013. Trade adjustment assistance for firms: Economic, program, and policy issues. Washington, DC: Congressional Research Service.Google Scholar
- Mundell, R. 1957. International trade and factor mobility. American Economic Review, 47(3): 321–335.Google Scholar
- Polanyi, K. 1944. The great transformation: The political and economic origins of our time. New York: Farrar & Rinehart.Google Scholar
- Rodrik, D. 2011. The globalization paradox: Democracy and the future of the world economy. New York: W.W. Norton.Google Scholar
- Rodrik, D. 2017. Straight talk on trade: Ideas for a sane world economy. Princeton: Princeton University Press.Google Scholar
- Rodrik, D. 2018. Populism and the economics of globalization. Journal of International Business Policy, 1(1).Google Scholar
- Schumpeter, J. A. 1942. Capitalism, socialism, and democracy. New York: Harper & Row.Google Scholar
- Slaughter, M. J. 2000. What are the results of product-price studies and what can we learn from their differences? In R. Feenstra (Ed), The impact of international trade on wages. Chicago: University of Chicago Press, 129–169.Google Scholar
- Tinbergen, J. 1954. International economic integration. Amsterdam: Elsevier.Google Scholar