At this particular moment, it is more than reasonable to wonder whether constitutional rights and human rights matter. Fortunately, a large number of legal scholars and political scientists have attempted to answer this question. Unfortunately, they have been using the wrong tools.
Over the past 10 years, there has been an explosion of what might be called “quantitative rights scholarship,” or attempts to measure, using quantitative instruments, whether rights “matter” or whether they have “failed.” This scholarship examines state commitments to basic rights—through domestic constitutions and international treaties—and asks whether states adhere to those commitments.
What does this data show? That depends on whom you ask. A wide array of scholars have looked at very similar data and drawn very different conclusions. In very general terms, there are two camps: One set of scholars argues that state commitments to rights have little measurable impact on state behavior (see, for example, Eric Posner and Adam Chilton and Mila Versteeg), while another set argues that state commitments to rights matter—but only under certain conditions and only after controlling for certain variables (see Chris Farris, Beth Simmons). To read this scholarship, one has the sense that rights either don’t matter at all or that they matter a little bit, sometimes.
But the two camps have more in common than their disagreement suggests: Both believe that the answer to the question “Do rights matter?” can be found by comparing state behavior to state commitments across large numbers of states.
I disagree, and I’ve written an essay explaining why. Large-N empirical studies can be enormously useful in many contexts, but in my view they are ill-suited to answering deeper questions about rights. As I say in the introduction:
Imagine asking, “does Catholicism matter?,” and then attempting to answer the question by measuring—with sophisticated quantitative tools—how the Vatican behaves towards Catholics. Not only is there a mismatch between the question and the empirical inquiry, but the question misses the point.
Part of the problem is methodological. These studies typically examine state behavior over a relatively short time span, whereas rights can take decades or centuries to mature. Indeed, if one were to deploy the same methodology used by the scholars of quantitative rights to study, say, the Bill of Rights in the 19th century, one might have concluded—somewhat prematurely—that the Bill of Rights did not matter. These quantitative studies have other defects as well. They typically rely on an overly restrictive dependent variable, as Ryan Goodman and Derek Jinks note. That is, they typically focus only on state action, though rights can of course be violated by nonstate actors as well. And perhaps most troublingly, these studies typically rely on quantitative indicators of well-being that give at best a weak measure of rights fulfillment. These quantitative indicators—data sets that attempt to score a state’s performance along some human rights measure—are notoriously flawed.
To understand just how flawed they are, consider China. Perhaps the most well-regarded data set for country-level human rights conditions is the CIRI data set (available here). Much of the quantitative rights scholarship looks to the CIRI data set to evaluate whether state scores change in the years after rights commitments. To see how well this data set captures conditions on the ground, here is a plot of the score that China receives on the CIRI data set across a wide range of human rights metrics for the past 30 years:
China has received the same score every year for the past 30 years, despite enormous gains in both civil and socioeconomic rights. This is beyond counterintuitive. As a result of deliberate state action, China has lifted hundreds of millions of people out of poverty—arguably the single greatest advancement of human welfare by a state in the last century—and these gains are ignored by some of the dominant human rights indicators.
But the problem is more fundamental than merely finding appropriate dependent and independent variables. To the extent that these studies say they show that rights “matter” or, alternatively, have “failed,” the studies claim more than they can deliver. It is not at all clear that rights can be reduced to such simple metrics. As I write in the essay:
Rights are more than just a promise to which the state can be held. Indeed, a case could be made that this is the least compelling aspect of a right. Rights can also be: a language for articulating grievances; a tool for building a social movement; a specific aspirational goal for the country (“we don’t want to condone that kind of thing”); a general aspirational goal (“we want to be the kind of place that doesn’t do things of that nature”); and more.
Rights are both more durable and more fragile than the current quantitative scholarship suggests. Take an example from this country’s recent past: The United States is committed to a right against torture; in 1994, it ratified the international Convention Against Torture. The Bush administration later engaged in behavior that seemed to violate this rights commitment. What does that conduct tell us about the right? To some quantitative rights scholars, it seems to suggest that the right against torture has failed.
But did it? The Bush administration’s harsh interrogation program was eventually rolled back amid a public outcry. Today, the anti-torture norm may well be stronger among interrogators than it was before the enhanced interrogation tactics were proposed during the Bush presidency. The full measure of a right cannot be captured in any meaningful sense by a single metric, and especially not state conduct toward citizens measured over a short timeframe.
That is good news for those wondering about the damage a Trump presidency might bring to our most sacred constitutional values. The fact that a leader behaves as if he does not care about rights—the rights of immigrants, the rights of women—is not proof of those rights withering. To the contrary, the language of the resistance movements is steeped in rights talk. If rights are a rallying cry for social movements and a language for politics, then they are alive and well in the Trump era. To Donald Trump and his allies, rights may not matter; the fact that his political enemies thrive on his disregard for rights is proof that they still matter a great deal.
This is not to say that rights talk is a helpful mode of political discourse, nor that today’s rights movements are problem-free. I try to catalogue some of those problems in the paper, including the way that rights discourse often wrongly frames hugely complex problems as questions of individual culpability. Others have explored the limits of human rights discourse in more detail—none more thoughtfully than Sam Moyn, whose recent book, “Not Enough,” is brilliant on this issue. Rather, the point is that this is the conversation worth having: how rights discourse functions as a source of political and legal power, and with what limitations. So far, that conversation has not benefited significantly from efforts to quantify rights promotion.