How We Stop Talking Past Each Other: A Rejoinder to Hoekman and Nelson’s Reply to My Article on Narratives about Winners and Losers from Globalization

When Donald Trump was elected to the US presidency, the instinctive reaction of many public officials, trade economists and international economic lawyers was to fight back. And fight back they did – in reports, op-eds, blog posts, and interviews. It did not appear as though it would be particularly hard to win the argument: Trump’s economic illiteracy was evident even to the casual observer, and his policy prescriptions were routinely exposed as ill-informed and misguided not just by trade experts, but also by mainstream journalists and even comedians. However, while ‘winning’ these arguments reassured those of us who understand the theory of comparative advantage and are aware of the existence of global value chains that we had the theory and data on our side, it did not facilitate a genuine dialogue with those who had embraced Trump’s vision of trade – or, for that matter, with voices on the left who had long been critical of trade agreements. What was required, it seemed to me, was an attempt to step back from the arguments and to take a bird’s eye view of the debate. This is what I tried to do in my article “How Should We Think about the Winners and Losers from Globalization? Three Narratives and Their Implications for the Redesign of International Economic Agreements” (published in the latest issue of EJIL 30 EJIL (2019) at 1359): I tried to provide a map that showed where the contending narratives overlap and where they diverge. In mapping the narratives in this way, I suspended judgement about the narratives’ empirical and normative merits.

While Hoekman and Nelson criticise the article on a number of grounds, it is fundamentally with this suspension of judgement that they take issue: they do not see the point of presenting the narratives without evaluating them against some sort of “intellectual benchmark” (30 EJIL (2019) at 1401); worse, they believe that doing so has “nihilistic” implications (Ibid. at 1408). In this rejoinder, I want to explain why I think that the type of analysis that I provide in the article is useful. I should say at the outset that, contrary to what Hoekman and Nelson claim, I do not believe – and I hope I say nothing to suggest – that “an effort to assess if the stories are empirically founded is not of any use” (ibid.). To the contrary, I believe that such an effort is valuable and an essential precondition for effective policymaking. However, many scholars and analysts are already engaged in that effort. In the article, I was trying to make a different kind of contribution that was largely missing in the discussion of trade policy: I was trying to encourage the participants in the debate to hear each other out and to stop talking past each other. Doing this effectively required me not to rush to judgement and not to adopt the standpoint of any of the narratives as my own. This is not to say that there is no place for applying intellectual benchmarks to trade policymaking, just that doing so in this article would have been premature. In fact, my hope was that the type of dialogue that I was trying to facilitate with the article would help lay the groundwork for the construction of a new intellectual benchmark for trade policymaking that would be responsive to the concerns that animate all three of the narratives (and potentially others).

What insights can we gain from reconstructing contending narratives about winners and losers from globalization, even without evaluating them? First, reconstructing a narrative allows us to understand the narrative’s internal logic, which can in turn help us make sense of and even predict the policy prescriptions that the narrative generates. Second, stepping into the narrative can give us an appreciation of the appeal of the narrative to those who embrace it. And finally, reconstructing a narrative enables us to identify not just the empirical claims, but also the normative judgements that are embodied in it.

Understanding the internal logic, appreciating the appeal and identifying the normative judgements of a narrative are, in turn, important preconditions for a (potentially) fruitful dialogue between the proponents of different narratives. When we articulate the logic of a narrative, we concede that it has a certain coherence and is not just crazy-talk – in other words, we signal that we take the proponent of the narrative seriously as an interlocutor. In Trump’s case, it is admittedly tempting to dismiss his possibly made-up stories about “massive” miners hugging him backstage and hyperbolic claims about “jobs theft” as some sort of bizarre ramblings. However, any type of engagement with proponents of a narrative requires us to distill the propositional content of their language and to identify the role that anecdotes such as these play in the larger story that Trump tells about trade. The trope about the “massive” miners, for instance, can be understood as an expression of Trump’s view that blue-collar workers cannot be expected to accept jobs that are inconsistent with their manhood (so-called “pink-collar” jobs), and the talk about jobs ‘theft’ makes sense as an allusion to the conception of jobs as being akin to property, with the implication being that a job loss is a wrong to be righted.

Understanding the logic of the narrative is a first step to appreciating its appeal. It is hard to engage in a dialogue with someone if we do not understand what motivates them and how they see the world. In other words, grasping the logic of a narrative opens the door to empathy – to a recognition of what makes one narrative attractive and renders another one off-putting. For example, if we do not dismiss Trump’s habit of talking about jobs as akin to pieces of personal property that can be “stolen” and “shipped away” as meaningless rhetoric, we can start to understand why his conception of job losses might resonate with blue-collar workers who feel that they have lost much more than a source of income. Moreover, by stepping into the shoes of a proponent of the Trump narrative, adherents of the establishment narrative can gain a new appreciation of their own narrative and can ask: what does the particular way in which we see the world miss? How do the euphemisms with which our narrative is replete – for example, when we refer to people’s inability or unwillingness to uproot themselves and their families as “labour mobility frictions” (as the World Trade Organization does here, at 16) – sound to those whose life experience we describe in those terms?

Finally, reconstructing the narrative lays bare not just the empirical assumptions, but also the normative building blocks of the narrative. The presence of these normative elements means that anyone who wants to convince proponents of another narrative of the ‘errors of their ways’ will not only have to marshal empirical evidence, but also normative arguments. And while we may be able to settle debates about empirical questions with the methods suggested by Nelson and Hoekman (assuming that the proponents of the other narrative are receptive to empirical arguments), the normative elements require a deeper engagement. In this process, the different narratives can again serve as mirrors for each other: the normative claims of one narrative highlight the normative choices implicit in other narratives. For instance, the categorical distinctions that the Trump narrative draws between different types of jobs, as well as its dismissal of the normative significance of consumption, draw attention to the ways in which the establishment narrative is agnostic as to who does what where and treats income gains derived from production (through wages) and consumption (through lower prices) as fungible.

The outcome of the normative engagement between the narratives is fundamentally open: some participants in the dialogue may end up rejecting the normative positions of other narratives, while others may incorporate some of them into their own narrative. One could imagine, for instance, that a normative engagement between the Trump narrative and the establishment narrative could lead to a new narrative that accounts for the non-monetary significance of jobs without privileging traditional manufacturing jobs that are predominantly held by men, in the way that the current version of the Trump narrative does (see my discussion of the gendered nature of the Trump narrative: 30 EJIL (2019) at 1367). Suffice it to say that, if we do not engage with the normative judgements embedded in each narrative, there is no reason to expect that we can develop an “intellectual benchmark” for the evaluation of trade policy that holds appeal to the adherents of more than one narrative.

We need look no further than Hoekman and Nelson’s rendering of the Trump narrative to see how difficult it is to grasp the logic, appeal and normative foundations of a narrative if we approach the analysis of that narrative with a preconceived “conceptual framework” (30 EJIL (2019) at 1401) in mind. Hoekman and Nelson do not engage with what Trump and other proponents of his narrative actually say. As a result, they misread Trump’s ‘jobs-as-property’ rhetoric as an expression of the lump of labor fallacy, even though there is no evidence that Trump or his supporters believe that there is a limited number of jobs. The Trumpian concern about jobs ‘theft’ stems from the fact that he and his supporters only care about specific types of jobs, namely manufacturing jobs, whose number is in fact limited. Moreover, Hoekman and Nelson ridicule the idea that Trump’s habit of talking about jobs as akin to property might resonate with workers by pointing out that ‘virtually no one, at least in a market economy, believes that people have a right to a specific job’ (30 EJIL (2019) at 1403). They are right in the sense that nobody argues that people literally have a right to a specific job; however, the very ‘essence of metaphor is understanding and experiencing one kind of thing in terms of another’ (G. Lakoff and M. Johnson, Metaphors We Live By (2003), at 5) without believing that the one thing is the other thing. Because Hoekman and Nelson dismiss the significance of Trump’s metaphorical language, they do not pursue the question of whether this language might appeal to workers because it captures their experience of job losses better than the language used by the other narratives.

Finally, Hoekman and Nelson refuse to concede that the differences between their framework for assessing the merits of trade policies and those of the other narratives are at least partly normative in nature. Specifically, they argue that there is ‘nothing “normative”’ (30 EJIL (2019) at 1405, fn 14) about treating the income effects of lower prices and higher wages as fungible, since both affect a household’s purchasing power in the same way. In their telling, even the “psychic losses” associated with unemployment can be factored into the overall welfare calculation, since they are “no more absolute than the effects that work through prices” (ibid.). In their view, all the value judgements that proponents of the Trump narrative make can be converted into “positive issue[s]” (ibid.) that are amenable to a calculus of welfare maximization. I do not deny that this can be done – and it may well give us comfort that we are pursuing the theoretically “best” policy. But dissolving all categorical distinctions in the blur of a single “welfare” metric is not only in itself a normative move, but also ignores how the people who are affected by the policies in question actually talk and think about their experience (Zachary Liscow makes a similar argument in his work on the “one-pieism” of law and economics scholarship). It therefore prevents us from responding in differentiated ways to the different dimensions of the losses that people suffer and of the injustices that they perceive.

In sum, while I do not question that we ultimately need an intellectual benchmark to design sound trade policies, I see a danger that approaching the analysis of narratives with a particular conceptual framework already in mind will hamper our ability to immerse ourselves in those narratives and to grasp their logic, appeal and normative elements. At a time when the foundations of trade policy are being fundamentally contested from various quarters, I believe that there is merit in suspending judgement – if only temporarily – in the interest of promoting a dialogue between the proponents of different narratives.

While I thus profoundly disagree with the thrust of Hoekman and Nelson’s critique, they make two observations that are well-taken and that I will incorporate in future work. I agree that my labels for the narratives are not ideal and that there are more narratives to be considered. Both the number of narratives that I discuss and the labels that I chose stem from the particular context that is the focus of the article, namely the debates about trade in the context of the NAFTA renegotiation, in which Trump and his allies were confronting the trade establishment, on the one hand, and were partially aligned with left-wing critics of trade agreements (hence the term “critical” narrative), on the other hand. In a forthcoming book that Anthea Roberts and I are currently working on, we rechristen the Trump narrative as the “protectionist” narrative and the critical narrative as the “corporate power” narrative. We also provide a more holistic picture of the contending narratives about winners and losers from globalization, including new narratives based on security and environmental concerns.